Has Political Devolution Led to Better Investment Returns? A Look at English Governance Mechanisms Over The Last Two Decades

Has Political Devolution Led to Better Investment Returns? A Look at English Governance Mechanisms Over The Last Two Decades | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Has Political Devolution Led to Better Investment Returns? A Look at English Governance Mechanisms Over The Last Two Decades Sanjay Raja This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8896200/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Despite the rapid expansion of devolution in England over the past two decades, with local andregional governments gaining unprecedented control over economic policy levers, the anticipatedeconomic dividends remain elusive. Using novel data on regional gross capital formation from 2000to 2019, this paper delivers one of the first quantitative assessments of the relationship between thescale of devolution and investment returns across key sectors. Contrary to prevailing policy narrativesand international evidence suggesting decentralisation boosts economic performance, our findingsreveal no significant link between increased devolution and improved investment returns in England.Robustness checks reinforce this outcome, highlighting persistent challenges such as limited fiscalautonomy, institutional capacity constraints, and ill-defined economic geographies. These results callinto question the effectiveness of England’s current devolution model in driving better economicprioritisation and underscore the urgent need for further research into the mechanisms required tounlock the full potential of place-based policy reforms. Devolution investment regional governance economic development decentralisation spatial policy United Kingdom Figures Figure 1 Figure 2 1. Introduction Devolution has emerged as a cornerstone of UK regional economic policy over the past decade, reshaping governance and offering a pathway to address spatial inequalities through empowered local decision-making (Warner, et al., 2024). But has devolution in the past two decades delivered better economic returns? From the City Deals to the Growth Deals, the last decade has seen an increase in decentralisation, with national governments seemingly shifting away from top-down economic development to bottom-up place-based policies (O'Brien & Pike, 2020). As part of England's devolution drive, local governments have taken back more control in transport, skills development, and housing, with city regions also able to control the economic prioritisation of investment projects. As of 2025, there were a total of eleven metro mayors in charge of combined authorities – a eleven-fold increase since 2000. Little has been done quantitatively to assess the impact of devolution in England on economic returns. With devolution seen as a solution for regional inequality, breaking the stronghold of centralised decision making, and allowing local leaders to tailor policies and investments to specific needs and opportunities, the need to unpack its effects (even at its early stages) has never been more important. Indeed, the empirical case for devolution enabling regions to act as policy laboratories, experimenting with new approaches and learning from each other to drive higher returns on investment and innovation via better resource allocation deserves more attention. The case for devolution is compelling: empowered localities make bold decisions that central governments avoid (GLA Economics, 2022). International evidence on devolution and its effects highlight some positives, with Burret et al. (2022) finding that competitive elements of decentralisation in Switzerland lead to improved economic performance, while Muringani et. al. (2019) show that decentralisation boosts regional GDP growth in the EU but only if paired with high-quality regional governance, as effective local institutions enable better public goods delivery and policy responsiveness. For such a policy relevant subject, this paper presents one of the first quantitative analyses looking at the impact of devolution (i.e. governance mechanisms) and investment returns in England, using novel and experimental data on regional gross capital formation between 2000 and 2019. Why investment returns as a measure of economic success? Simply, because devolution (i.e. the transfer of powers from central government to local government), in theory, should lead to better investment returns, both public and private, by enabling more responsive, place-based decision-making and reducing bureaucratic barriers for investment and growth. This can be achieved via better local knowledge and strategic alignment, integrated planning and infrastructure settlements, enhanced coordination, and most importantly local ownership of economic development. The paper looks to add to the empirical literature focusing on the returns to devolution, particularly at a time when focus on regional policy has increased significantly. I find no relationship between devolution and investment returns across the full range of categories. Put simply, I find no evidence that devolution has led to better decision making when it comes to the economic prioritisation of projects – something that sits at odds with the UK’s aim to deliver better economic prioritisation of investment via devolution. What does this suggest? Unlike other studies looking at broader European trends highlighting positive investment returns from better governance (Crescenzi, et al., 2016) something about devolution in England seems to be not working – at least as far as these mechanisms go. Indeed, in focusing on changes in regional governance mechanisms, least at its initial stages, we are yet to find any strong success stories coming from the recent devolution drive. And while there may be good reasons why more devolved regions have yet to show – at least statistically – any positive and significant returns, relative to less devolved or decentralised government mechanisms in place, our findings seem to chime with some of the long-standing issues prevalent when looking at English devolution. The lack of fiscal decentralization (i.e. revenue powers), institutional capacity, and poorly defined economic functional areas may indeed be limiting the potential positive effects traditionally seen via regional devolution. As focus on UK productivity and the lingering productivity puzzle increases, more research will need to be undertaken to investigate what drivers may ultimately be needed to better devolution outcomes in relation to investment returns in England. This paper is structured as follows. The following section sets out the context around devolution, investment and economic growth, whilst also briefly outlining the state of regional devolution in England, and how place-based policy focused on the prioritisation of investment projects in the last decade. The next section presents the empirical approach in estimating the economic returns of investment associated with the scale of devolution in England, alongside its limitations. Following this, the section concludes with descriptive analyses, baseline estimation results, and a series of robustness checks. The last section presents the conclusions while discussing the broader implications of the study. 2. Devolution, Investment, Economic Growth a. Devolution and Economic Growth The term ‘devolution’ has often been reflected the decentralisation of resources from central governments to local governments. This, however, represents a narrow definition. As Rodriguez-Pose and Gill (2004) highlight, devolution “can assume various forms, ranging from decentralisation of power and legitimacy through to a mere delegation of responsibilities and financial duties”. The link between political devolution and economic growth is complex with much of the focus on devolution based on fiscal decentralisation as opposed to political governance changes. Decentralization fosters policy experimentation and competition, as regions adopt best practices from peers ("laboratory federalism"). Tiebout (1956), for example, posited that decentralization fosters competition among regions, enhancing growth. The model argues that competition among local governments replaces political processes in achieving efficient public goods allocation. Individuals "vote with their feet" by relocating to jurisdictions offering their preferred mix of taxes and services, eliminating the need for centralized preference aggregation. Zhao et. al (2021) found that fiscal decentralization significantly promoted regional innovation efficiency in China, with both vertical competition (between central and local governments) and horizontal competition (among local governments) further enhancing regional innovation efficiency. The authors conlcuded that local governments with more fiscal autonomy can better target investments and policies to stimulate innovation, creating positive externalities for innovation participants. Rodriguez-Pose and Ezcurra (2010), however, showed that any positive effects coming fiscal decentralisation or competition were highly state-dependent. As part of their empirical analyses, they concluded that decentralization was not systematically associated with changes in regional disparities – whereby in high-income developed countries, decentralization was generally linked to a reduction in regional inequality, suggesting that devolved powers could help even out regional growth when institutional capacity and redistributive mechanisms were strong. In low and middle income countries fiscal decentralization was associated with a significant increase in regional disparities. Here, poorer regions often lacked the institutional capacity, tax base, and infrastructure to benefit from decentralization, leading to widening gaps. Elsewhere, Meloche et. al. (2004) examined the importance of fiscal autonomy in evaluating the effects of decentralization. Using data from European transition countries they found that fiscal autonomy—where subnational governments have control over their own revenues - was positively associated with economic growth. The study distinguished between decentralisation (the devolution of functions) and true fiscal autonomy, finding the latter to be more growth-enhancing. A separate study by the OECD (Blöchliger & Égert, 2013) analysed the relationship between fiscal decentralisation and economic activity concluding that fiscal decentralisation was positively and significantly related to GDP per capita, productivity, and human capital. They deemed that revenue-based decentralisation (tax autonomy) yielded stronger positive effects than spending-based measures. However, the relationship was non-linear, with diminishing returns as decentralisation increased, while intergovernmental transfers while negatively associated with growth. While shifting to political devolution mechanisms, however, the literature has centred around improved efficiency in public sector delivery as well as increased accountability/less corruptibility, allowing regions to better enact place-based policies to drive growth and foster more efficient investment outcomes. Oates (1973) via the “decentralisation theorem” posited that local governments were better able to match public goods to heterogeneous preferences, supported by welfare gains in heterogeneous jurisdictions. Oates argued that subnational governments could better tailor policies to local preferences, improving efficiency. Bardhan & Mookherjee (2006) concluded that decentralisation shifted accountability closer to citizens but risked capture where some locals could disproportionately influence service allocation. Hooghe et al. (2010) showed that regions with high levels of devolution (looking at institutional depth, fiscal autonomy, and law-making powers) were better able to align services like education and healthcare with local needs. A separate paper from the International Monetary Fund (IMF) highlighted that fiscal decentralization could lead to improvements in public sector delivery efficiency - but only under specific conditions, including where regions/localities had adequate political and institutional environments, a minimum degree of spending decentralization and a suitable degree of revenue decentralization (Sow & Razafimahefa, 2015). Evidence from Nordic countries, for example, show that municipalities with control of healthcare and education achieved better outcomes, including satisfaction (Hitchcock, et al., 2017). Faguet and Pöschl (2015) found that decentralisation in Bolivia improved public service alignment with local preferences, enhancing access and quality in education, healthcare, and infrastructure. In the UK specifically, Britteon et al. (2024) concluded that the Greater Manchester devolution deal lead to improved population health outcomes, driven by improvements in health services and wider social health determinants. Moreover, local governments have better information about their respective region’s needs and priorities and ultimately can be held more accountable to its electorate. Voters can reward/punish local leaders based on service quality. Faguet (2021) noted that decentralisation brings government "closer to the people," shortening and tightening the loop of accountability. Local officials become directly accountable to citizens, whose votes and voices become more powerful in influencing local service delivery and policy outcomes. Hirschman (1970) described this as an “exit and voice” mechanism, whereby citizens can monitor the quantity and quality of services, and discipline politicians by rewarding or sanctioning them in competitive elections. The empirical research surrounding this question has broadly confirmed the relationship between decentralisation/devolution, governance quality, and enhanced accountability. For instance, Fisman and Gatti (2002) examined the effects of decentralization and corruption via a cross-country analysis, concluding that “fiscal decentralization in government expenditure is strongly and significantly associated with lower corruption”. Bardhan and Mookherjee (2006) noted that decentralization shortened the chain between citizens (principals) and officials (agents), making it easier to monitor and sanction corrupt behaviour. The OECD also showed that decentralization can improve service delivery and accountability when paired with strong local institutions, transparency, and social capital (OECD, 2019). Others have also argued that federal systems with strong local oversight (e.g., Germany, Canada) exhibit lower corruption than highly centralized ones. b. Devolution in England – a brief look back at the last two decades Devolution and regional governance in England have a complex and fragmented history. The idea of English regional devolution dates back to the early 20th century, with Winston Churchill advocating for regional parliaments in 1912 (Torrance, 2019). Post-war reconstruction, guided by the Beveridge Report (Beveridge, 1942) and Keynesian economics, saw national government directly planning housing, welfare, and economic development. The National Health Service Act 1946 transferred health services from local governments to a centrally run NHS under the Ministry of Health. After World War II, several commissions, notably the Redcliffe-Maud Report (1969), recommended replacing England’s two-tier local government with larger unitary authorities grouped into eight provinces with devolved powers. However, these proposals were not implemented, as political focus shifted elsewhere. In 1994, the Major government established ten Government Office Regions to coordinate central government activities regionally. The Blair government (1997–2010) advanced regionalism by creating Regional Development Agencies (RDAs) in 1998 to foster economic growth, and indirectly elected Regional Assemblies or Chambers to oversee them. However, these assemblies lacked direct democratic legitimacy. A key moment came in 2004, when a referendum in North East England overwhelmingly rejected a proposed elected regional assembly. This defeat led to the abandonment of plans for further regional assemblies, and between 2008 and 2010, the existing assemblies and RDAs were abolished. The one major devolution success, however, came via the capital. The Greater London Authority (GLA), comprising a directly elected Mayor and Assembly, was established in 2000, marking the only successful city-regional devolution in England. Most councils until this point relied on a ‘two-tier government’ as per a ‘leader-cabinet model’ where councillors were elected and subsequently elected a leader. Since 2010, the focus for national governments shifted to “City Deals”, “Growth Deals”, and more-recently, the “Levelling-Up” agenda. The push towards devolution took the form of combined authorities and directly elected metro mayors. Combined authorities enabled groups of local councils to voluntarily collaborate, often led by directly elected mayors (e.g., Greater Manchester, West Midlands). These combined authorities gained devolved powers over transport, housing, and economic development. Some combined authorities even received increased fiscal decentralisation via a full scale (pilot) 100% business rates retention scheme (including Greater Manchester, Liverpool, West of England and the West Midlands). Devolution deals towards the end of the decade also included commitments to multi-year funding for specific projects, though recent waves of devolution have invoked some criticism. The Industrial Strategy Council recently noted that policies regarding sub-national governance “have been piecemeal rather than coherent strategies”, resulting in a “current local institutional landscape [that] is confusing and fragmented”, having “no clear, long-term, coherent vision for sub-national policy”. c. English devolution and economic returns of investment Economic policies need to be tailored to their local needs, given their respective industrial bases, demographics, human capital, and local expertise. In theory, local governments will almost always be better able to identify and prioritise economic projects – especially in the context of constrained and limited funding. This was the goal for better and deeper political devolution in England over the last several decades – and in particular, over the course of the last two decades where governments (past and present) sought to push local intervention in the prioritisation of economic projects in a bid to ramp up regional productivity convergence and growth. The proliferation of so-called ‘competitive bids’ and ‘spending pots’ required local governments (either via Local Enterprise Partnerships, City Deals, Growth Deals, and or Trailblazer Deals) to bid or make applications to central government to receive grant funding – where funding would be determined on the quality of potential economic returns associated with local projects and initiatives. Between 2010 and 2024, economic prioritisation of investment projects have come via a number of place-based policy schemes in an effort to maximise impact and improve economic returns (UK Government, 2015) via Local Enterprise Partnerships (LEPs), City and Growth Deals, and Enterprise Zones. The importance of political decentralisation combined with local authorities will almost always be better able to channel local expertise and knowledge as well as fostering the ‘right’ type of investment – and most importantly, investment projects that align with each region’s local strategic. The capability to better identify and prioritise high-return projects, target investment into sectors with regional comparative advantage, and respond more nimbly to local needs can boost investment returns in not just the public sector, but the private sector too. Furthermore, the ability to strike multi-year funding arrangements can also help ‘crowd-in’ private sector investment in contribution to various regions’ local ambitions. The introduction of long-term planning can reduce investment risk more broadly across regions for both public and private actors. Integrated strategies for transport, housing, and skills are critical for unlocking private sector investment and improving public investment efficiency. The fostering of strong partnerships between public bodies and private investors could enable co-investment and leveraging public funds to crowd in private capital, making regions more attractive to investors. 3. Empirical Model and Regression Estimations a. Model Specification and Data The empirical analysis aims to shed any insight on whether political devolution helped shape investment returns across England given the implicit focus around the economic prioritisation of investment projects. The analysis relies on multiple channels of investment using novel and experimental investment data from the Office for National Statistics (ONS). Including various categories of investment (i.e. gross fixed capital formation) allows for an investigation that reflects structural differences in specific regional investment categories, allowing for a more thorough investigation of various investment decisions based on regional/local political preferences, the financial requirement to deliver them, and consequently, the potential impact on regional economic growth. The impact of devolution is modelled by including an interaction term between proxies for investment and the scale of English devolution. The following model is used to estimate the link between devolution and investment returns on economic growth: Using novel and experimental data from the ONS, the paper utilizes the following breakdown of gross capital fixed formation as proxies for regional investment: Buildings and Structures : This includes everything from dwellings, roads and any other construction of buildings and structures; Transport Equipment : This includes ships, road transport vehicles, and all other transport equipment (including aircrafts); ICT Equipment : This includes computer hardware and telecoms; ‘Other Tangible Assets’ : This includes machinery and equipment, cultivated assets, and weapon systems; And Intangible Assets : This includes research and development (R&D), mineral exploration and evaluation, computer software and databases (both purchased and own account software), as well as entertainment, literary or artistic originals. All of the above investment proxies are expressed in the form of a natural logarithm and are normalised using regional population data (i.e. investment per capita). All data are deflated using relevant regional deflators. Turning to the political devolution proxy, and what is a new contribution to the literature, a political devolution variable is constructed based on a numerical score of 1-5 between 2004-2019 with 1 representing the lowest form of political devolution and 5 the highest form of political devolution. The scores assigned are attributed as follows: 1 = Two-Tier Local Council – the lowest form of devolution. The Two-Tier Local Council has decision making shared between the two tiers of local government (upper tier being county councils, and lower tier being district and borough councils). 2 = Cabinet System. This is a small group of elected councillors, where power is limited but concentrated in the cabinet. 3 = Unitary Authority. This is where a single council handles all local government functions (replacing the two-tier system). 4 = Combined Authority. This includes multiple local authorities to oversee regional issues, with transfer of certain powers from central government to the combined authority enabling regional decision-making. 5 = Elected Mayor – the highest form of devolution. This represents a directly elected mayor that leads the local authority or combined authority, providing leadership and accountability for local and regional governance. The choice of political devolution correspond to meaningful structural reforms in governance, with each level representing incremental increases in both legal and decision-making powers transferred from central government to local government (see (Local Government Association, 2024)). Two-tier local councils and cabinet-style systems, for instance, have limited control over basic services, planning and regulatory functions. Unitary authorities and combined authorities have greater control over major public services, economic development, and infrastructure. While elected metro-mayors provide executive leadership and accountability with increased discretion over locally-raised and devolved funds, policy priorities, and regional coordination. Combined authorities as well as elected metro mayors see additional fiscal autonomy too, with powers over transport, skills, housing, adult education, as well as – in some cases – increased ability around tax retention. To understand how devolution affects investment, an interaction of the devolution variable and investment is used. The spatial level at which the model specification is run on is the International Territorial Level 3 (ITL3) level yielding a more granular - yet economically relevant - level of spatial analysis on devolution and economic returns. The vector of controls includes a number of variables traditionally linked to impacting economic growth. In accordance with the endogenous growth theory, we include proxies for sectoral composition, agglomeration effects, human capital, jobs density, spillovers of regional investment. The sectoral composition is defined as the GVA share for production, services (excluding public services), and public services. Agglomeration effects are proxied for by population density – defined formally as population per square kilometre (see Ciccone & Hall, 1996, Glaeser & Resseger, 2010). Human capital is proxied by the share of the regional population with NVQ4+ education (Benhabib & Spiegel, 1994, Hanushek, E. A., & Kimko, 2000, Gennaioli, et al., 2013). Jobs density acts as a proxy for labour demand – another important determinant in driving economic growth (Acemoglu, 2002, Autor, , et al., 2003), and is defined as the number of jobs per working age population (those aged 16-64). Given that investment in buildings/structures, transport, ICT, tangible assets and intangible assets will likely impact economic growth and performance beyond the specific area in which the investment takes place, we also account for any spillovers generated from investment with a spatial lag. Estimation of the spatial lag is based on an exponential decay function, similar to the one used by Raja and Larsson (2024). To deal with the changing nature of geographical data in the UK, a uniquely constructed ITL3 dataset is used, comprising of 114 regional areas. For all 114 ITL3 areas, there is a fully balanced dataset. The time period used in the baseline estimations is 2000-2019 – providing nearly two decades of data. b. Descriptive Statistics Table 1 presents a summary of the key statistics included in the dataset, highlighting both the mean and standard deviation of key variables used in the econometric specification in years 2000 and 2019 for comparison. Looking across all 114 ITL3 regions constructed for the analysis, average regional GVA grew by roughly 2% per annum between 2000 and 2019. The broad sectoral composition of the economy also shifted as one would expect, with the share of production across all ITL3 regions narrowing by 2 percentage points. The services share across all ITL3 regions increased by 5 percentage points. And the public sector share of all ITL3 regions dropped by 2 percentage points. Looking at investment levels, one can see similar increases in investment levels (GFCF), with total gross fixed capital formation levels growing by roughly 2% per annum across the time sample. Looking at the disaggregated breakdown of investment, buildings growth saw the largest growth at 3.8% per annum. Intangibles investment growth averaged nearly 3% per annum. ‘Other tangibles’ and intangibles investment barely grew. And ICT investment contracted between 2000-2019, reflecting potential falls in cost over time (see Coyle & Hampton, 2023) (Coyle & Hampton, 2023). Other included variables track generally as expected over the last two decades. Jobs density (i.e. the number of jobs per working age population), for example, increased by 0.07pts. The share of the population with NVQ 4+ qualifications increased by 17 percentage points over the same time period. And the population density across ITL3 regions increased by nearly 400 people per square kilometre. Table 1: Descriptive Statistics 2000 2019 Observations Mean Std. Deviation Observations Mean Std. Deviation GVA (£, millions, real) 114 10,690 8,629 114 15,056 14,502 GVA, Production Share 114 17% 8% 114 15% 8% GVA, Services Share (excl public servcies) 114 53% 11% 114 58% 10% GVA, Public Services Share 114 23% 7% 114 21% 6% Jobs Density 114 0.83 0.49 114 0.90 0.56 Skills Share (NVQ 4+) 114 23% 7% 114 39% 11% Population Density 114 2,205 2,521 114 2,601 3,116 GFCF Total Investment (£, millions, real) 114 2,133 1,597 114 3,054 2,590 GFCF Buildings Investment (£, millions, real) 114 1,007 661 114 1,734 1,394 GFCF Intangibles Investment (£, millions, real) 114 441 493 114 685 913 GFCF Other Tangibles Investment (£, millions, real) 114 337 290 114 374 371 GFCF ICT Investment (£, millions, real) 114 209 179 114 116 123 GFCF Transport Investment (£, millions, real) 114 140 103 114 145 147 Source: Office for National Statistics, NOMIS Turning to the devolution variable, Table 2 presents a summary table of the five basic categories of political and economic devolution used in the analysis to depict devolution governance mechanisms across England: two-tier local councils, cabinet system, unitary authorities, combined authorities, and elected mayors. Since 2010, there has been a marked increase in devolution with several ITL3 regions falling into either a combined authority or metro-elected mayor governance mechanism. In 2019, 56 regions are identified as part of either a combined authority or part of a metro-elected mayoralty – three times as many as in 2010. Figure 1 presents a spatial visual of the assigned devolution scores across England. These are based on the final devolution mechanisms in place for ITL3 regions as of 2019 – the final year of the analysis. Figure 2 presents a scatter-plot of regional devolution scores at the ITL3 level, relative to the cumulative investment (GFCF) growth between 2000-2019. A visual inspection of the data suggests no obvious relationship between devolution level and investment growth but we do see those ITL3 regions at the top end of the devolution spectrum registering the largest cumulative increase in investment growth. Table 2: Governance Mechanism By Year Governance Mechanism, # of ITL3 Regions 2010 2019 2024 Two-Tier Local Council 34 23 17 Cabinet System 19 3 3 Unitary Authority 39 32 37 Combined Authority 0 8 3 Metro Elected Mayor 22 48 54 Source: Author’s calculations c. Estimation Issues and Specifications The empirical model outlined in Equation (1) is estimated using fixed effects panel methods, incorporating time dummies to account for temporal variations. To address potential issues of serial correlation and heteroscedasticity, clustered standard errors are employed (at the ITL3 level). Additionally, the impact of spatial autocorrelation—where error terms of neighbouring observations may not be independent—is mitigated by including spatially lagged variables as controls. These variables explicitly capture interactions between neighbouring regions, thereby reducing their influence on the residuals. Using fixed effects in a panel regression controls for unobserved and time-invariant heterogeneity across all ITL3 regions, ensuring that the estimated coefficients are not biased by omitted variables that differ across entities but remain constant over time. Including time effects accounts for common temporal shocks or trends that affect all entities simultaneously, isolating the impact of variables of interest from broader time-specific influences. Together, fixed effects and time effects enhance the robustness and accuracy of the regression results by addressing both cross-sectional and temporal variations. The regression specification also tries to mitigate against omitted variable bias through the inclusion of temporal lags – in this case, GVA per capita. Lagged GDP/GVA growth can act as a proxy for unobserved factors that affect current growth, such as institutional quality, cultural factors, or historical events. Moreover, given that GDP/GVA growth rates are often serially correlated, including lags helps account for this autocorrelation, improving the efficiency of the estimates (see Beck and Katz, 1995; Arellano & Bond, 1991). The baseline estimations are provided in Table 3. Column 1 includes a stripped down version of the specification, including only the lagged GVA growth of a region and the devolution variable. Column 2 introduces total gross fixed capital formation investment growth. Columns 3-7 include disaggregated investment categories. In columns 2-7, the regression specification includes the devolution proxy, investment growth, and the interaction between the devolution variable and investment growth. The interaction term tests whether the effect of investment growth on GDP growth differs in regions with varying levels of devolution. Put differently, the specification tests whether investment returns are stronger as a result of increasing devolution. d. Baseline Estimations In all specifications, the devolution variable is positive, but not a statistically significant factor when it comes to driving GVA per capita growth. Changes in investment (total, intangibles, buildings, other tangibles, ICT, transport) are all positively linked to GVA per capita growth – but are not statistically significant. The key devolution variable interacted with changes in investment (be it total investment or disaggregated investment) is also not statistically significant in any of the investment specifications included– and in some cases (i.e. total and transport) are negatively signed. What variables show up as statistically significant? Lagged GVA per capita – a sign of persistence is negatively signed but always statistically significant. Spatial spillovers of investment (intangibles, other tangibles, ICT) have a statistically significant relationship with GVA per capita growth. Industry shares also show up as statistically significant. Table 3: Baseline Regression Results Dependent Variable: Change of Log GVA Per Capita Total Intangibles Buildings Other Tangibles ICT Transport (1) (2) (3) (4) (5) (6) (7) Lagged GVA Per Capita -0.073*** -0.182*** -0.181*** -0.182*** -0.182*** -0.182*** -0.182*** (0.0167) (0.0248) (0.0247) (0.0248) (0.0249) (0.0250) (0.0258) Devolution -0.000515 9.37e-05 -5.00e-05 6.28e-05 5.69e-05 6.45e-05 0.000103 (0.00103) (0.000946) (0.000951) (0.000951) (0.000944) (0.000945) (0.000955) Change Investment (GFCF) Per Capita 0.00759 0.00213 0.000536 -0.000270 0.00122 0.00473 (0.00527) (0.00450) (0.00416) (0.00304) (0.00241) (0.00410) Change Investment Per Capita x Devolution -0.000425 0.00170 0.000324 0.000800 0.000278 -0.000248 (0.00172) (0.00171) (0.00136) (0.000959) (0.000780) (0.000878) Change in Spatial Weight of Investment Per Capita -0.000112 0.00505** -0.00206 0.00316** 0.00285** 0.00102 (0.00303) (0.00221) (0.00232) (0.00153) (0.00143) (0.00144) Production Share (GVA) 1.88e-05*** 1.87e-05*** 1.90e-05*** 1.92e-05*** 1.91e-05*** 1.92e-05*** (3.54e-06) (3.52e-06) (3.53e-06) (3.55e-06) (3.55e-06) (3.66e-06) Services Share (GVA) 4.51e-06*** 4.52e-06*** 4.55e-06*** 4.48e-06*** 4.49e-06*** 4.35e-06*** (8.38e-07) (8.32e-07) (8.42e-07) (8.52e-07) (8.53e-07) (9.21e-07) Public Sector Share (GVA) 7.06e-06 7.04e-06 6.72e-06 7.05e-06 7.06e-06 8.38e-06 (5.88e-06) (5.92e-06) (5.92e-06) (5.86e-06) (5.87e-06) (6.07e-06) Population Density -1.01e-05 -1.03e-05 -1.01e-05 -8.01e-06 -7.98e-06 -9.90e-06 (7.14e-06) (7.02e-06) (7.13e-06) (6.67e-06) (6.70e-06) (7.58e-06) Jobs Density -0.0190 -0.0200 -0.0188 -0.0182 -0.0182 -0.0174 (0.0187) (0.0187) (0.0188) (0.0190) (0.0190) (0.0189) Human Capital 0.0278 0.0281 0.0274 0.0218 0.0216 0.0176 (0.0352) (0.0350) (0.0351) (0.0350) (0.0350) (0.0374) Constant -0.251*** -0.705*** -0.701*** -0.707*** -0.710*** -0.710*** -0.708*** (0.0622) (0.103) (0.103) (0.103) (0.104) (0.104) (0.109) Observations 2,280 2,280 2,280 2,280 2,280 2,280 2,280 R-squared 0.613 0.471 0.473 0.470 0.472 0.472 0.480 Year FE YES YES YES YES YES YES YES Number of ITL3 Regions 114 114 114 114 114 114 114 Robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1 The insignificant relationship between the devolution variable, investment changes, and the interaction of devolution variable and investment (i.e. investment returns) all suggest that the link between the devolution variable and investment returns is less clear cut. There could be several reasons for this. For one, identification of such a relationship across 114 unique ITL3 regions may be complex and difficult to determine. As Cresenzi et al. (2016) noted, optimal investment would lead to zero marginal returns, given the potentially efficient allocation of resources, leading to a potentially weak statistical relationship between devolution and investment returns on paper. Diminishing marginal returns to capital may also play a role in the weaker statistical relationship found between the key variables of interest – though given the UK’s productivity problems, this seems less likely. What are other possible conclusions? Easterly (2002) cite that poor institutions could lead to capital misallocation, reducing any returns to investment. This would suggest that the link between deeper devolution and governance quality/institutional capacity may be weak – with regional institutions in the UK potentially lacking the necessary institutional structures and effective governance mechanisms to deliver better investment returns. Another potential explanation lies with the notion of poor management practices. Bloom et al (2013) show that poor management practices lower returns. Another potential explanation could also be with regards to the decision making processes within English regions. The introduction of deeper devolution to England across the country could have led to poor economic prioritisation, and in some cases leading to myopic investment decisions – decisions that may highlight important economic importance, but in reality do little to actually lift economic growth and/or productivity, see Laverty (1996) and Rodríguez-Pose & Gill (2020). Tomaney (2016) also noted that devolution in the UK did not show an equal shift in democratic advances with recent devolution programmes “embedded within a centrally imposed tax cutting agenda” lacking “fiscal solidarity and a genuine attempt to extend democratic accountability”. e. Robustness Checks In this section, several robustness checks are deployed to assess whether the above findings hold. In particular, the robustness checks employed include: a re-definition of the devolution variable, changing of the empirical time specification, change in dynamic specification, using alternative proxies for investment, and dealing with potential endogeneity concerns. Each are taken in turn below with results presented in Tables A1-A5. Redefining the devolution variable . One check to assess whether the above findings hold is by re-defining the devolution variable into a binary variable. Here only the two highest forms of devolution are assessed: combined authorities and elected mayors. In re-defining the devolution variable to only include the highest forms of devolution, measurement error and heterogeneity bias across governance mechanisms is minimised. Pooling all devolved regions together (including those with weak powers) could dilute the estimated ‘treatment effect’, leading to attenuation bias (measurement error) or spurious null results. Restricting the analysis, therefore, to ITL3 regions with meaningful devolution (including those combined authorities that saw increased fiscal decentralisation via 100% business retentions including West Midlands, Greater Manchester, and Liverpool) would ensure that the devolution variable defined in the empirical estimations is strong enough to detect any effect. There is empirical support to such a claim. Rodriquez-Pose and Ezcurra (2011) show large heterogeneity when looking at the effects of fiscal decentralisation and economic growth. Triesman (2007) showed that shallow devolution (e.g., administrative delegation) had no measurable economic impact, creating noise in pooled estimates. Furthermore, Enikolopov & Zhuravskaya (2007) showed that only strong devolution (with electoral accountability) improved governance and growth, while weak decentralization worsened economic outcomes. In re-estimating the baseline specification with the new devolution variable, the findings do not change – see Table A.1. The devolution variable, which is now a binary variable, is positive, but remains statistically insignificant. The key interaction variable in the empirical estimation shows mixed signs depending on the type of investment specified but crucially also remains statistically insignificant – reaffirming the findings in our baseline estimations. Changing the time specification. Another robustness check employed to validate the baseline estimation findings is by re-defining the time specification used for the empirical analysis. Here, instead of using the full dataset (2000-2019), a shorter timeframe is used. Specifically, the baseline estimations are re-run using a shorter timeframe of 2010-2019 – the period in which most devolution agreements in our dataset were agreed. This would also remove the period in which the global financial crisis occurred. Similar to the baseline estimations, both the devolution variable and the interaction between the devolution and change in respective investment variable remain statistically insignificant (see Table A.2). In fact, in all six specifications, the key devolution interaction variable with investment change is negative. Change in dynamic specification. The baseline model specified is a dynamic specification where the lagged level of GDP is used as a regressor. This, in effect, attempts to account for convergence as well as controlling for initial economic/GDP conditions of a region. As a further robustness check, a parsimonious version of the model is used, whereby the lagged GVA of a region is excluded. The results are presented in Table A.3. Again, the key interaction variable between devolution and change in investment remains statistically insignificant. The devolution variable is negative regardless of investment type specified. Alternative proxies for investment: road investment. In a similar analyses conducted by Crescenzi et al. (2016), the baseline equation is re-estimated using road investment data – instead of regionally disaggregated investment data. In England, the level of autonomy local and combined authorities have over transport projects depends on the specific devolution deal agreed with central government. Mayoral Combined Authorities, for instance, created under the Cities and Local Government Devolution Act 2016, often receive the widest set of powers. These can include control over bus services through franchising (as in Greater Manchester and the West Midlands), responsibility for developing and delivering a local transport plan, management of key road networks, and influence over local rail services in partnership with the Department for Transport. They may also receive devolved funding streams—such as the Transforming Cities Fund or City Region Sustainable Transport Settlements—allowing them to set long-term investment priorities without seeking central government approval for every project. Non-mayoral combined authorities and individual local authorities generally have narrower transport powers, often focused on local road maintenance, parking, cycling and walking schemes, and some bus service support, but they typically lack full control over public transport integration or multi-year transport budgets. As a proxy for infrastructure investment, the re-estimated model focuses on the following transport investments at the ITL3 level: total road length, motorways, A-roads, minor roads, trunk roads, and principle roads. Why look at transport investment in particular? Good transport infrastructure enhances local accessibility, and leads to increased activity via a reduction in transport costs and increased productivity (see Aschauer, 1989, Banister & Berechman, 2003, and Redding & Turner, 2015). As Crain and Oakley, 1995; Henisz, 2002; Acemoglu and Dell, 2010 highlight, local institutions – and by extension, regional governance mechanisms – that drive the state of incentives and constraints are effectively moulded by local institutions and the political capital of regions to maximise the returns to transport investment (Crain & Oakley, 1995, Henisz,, 2002, Acemoglu, & Dell, 2010). As Crescenzi et al. (2016) argues, “[inadequate] political institutions may negatively affect the economic returns to transport infrastructure investment well before the money is actually spent. Governments are directly responsible for appropriate infrastructure planning and rigorous project selection, making transport infrastructure planning and financing fundamentally a political topic.” As such, in using transport road data as proxies for transport investment, this particular robustness check assesses whether the earlier findings hold for a different proxy of investment. Results of the regression results are outlined in Table A4. Consistent with above findings, no statistically significant relationship is uncovered when looking at both the devolution variable and the interaction variable between devolution and change in transport investment. What does this suggest? Even when altering the investment variable, the model specification shows no statistically significant correlation between devolution and investment returns associated with transport/road investment, implying very little impact from English devolution on the economic prioritisation of transport investment – consistent with the baseline results. Dealing with endogeneity: a system GMM model. The estimated effect of devolution and changes in investment (proxied by various investment categories within gross fixed capital formation) on economic growth may be imprecise or biased if the direction of causality runs opposite to that assumed in the model - that is, if devolution and investment shifts are consequences, rather than causes, of regional economic growth. A large body of literature has sought to address the potential endogeneity of political decentralization and capital formation dynamics using instrumental variables. The model specified in Equation (1) includes two key variables—devolution and investment changes—that may be endogenous to growth, along with their interaction term, making identification strategies based on external instruments difficult to implement. Historical data on English regions, at the ITL3 level, are also not easily attainable. To minimise endogeneity concerns, one approach would be to employ a dynamic panel analysis using a GMM-system estimator – similar to that used by Calderon and Serven, 2004 and Crescenzi and Rodrıguez-Pose, 2012. Adopting a system-GMM estimator allows for a dynamic setup in controlling for unobserved heterogeneity and simultaneity while accommodating persistent growth effects. Put differently, this is a pragmatic approach used to circumvent the “instrument hunt” by relying on temporal dynamics. The re-estimated model specified includes the two main variables of estimation: the devolution variable and change in investment proxy interacted with the devolution variable. Due to the number of instruments, the estimated model is limited to including only lags in GVA growth, changes in investment, the change in spatial investment lag, skills share, jobs density and population density. Industry shares are therefore excluded in the GMM estimation. In order to maintain a lower number of instruments than regions, second-order time lags are used as instruments. The findings are in line with the baseline estimations. As per the GMM-system estimates, the devolution variable is consistently positive, but remains statistically insignificant. The interaction between devolution and changes in investment (proxied by total GFCF, intangible investment, buildings, ‘other tangibles’, ICT, and transport) show a positive relationship with GVA growth, but are also statistically insignificant. This is not to say that such a methodology solves for endogeneity concerns, however. Devolution remains an exogenous policy measure, and the effectiveness of devolution is largely path-dependent, where time-lags may not necessarily allow for exogenous variation (Bun & Windmeijer, 2010). At best, employing such a check allows for validation of the base model estimations. 4. Conclusions UK governments, past and present, have prioritised devolution as a means to narrowing the productivity gap between the north and the south of the country, while allowing regions to have a bigger say in the prioritisation of the economic projects, initiatives, and objectives of respective city regions. Indeed, from the City Deals in the 2010s to the Growth Deals and 'levelling-up' agenda that followed to the present version of devolution strategy (see Ministry of Housing, Communities, & Local Government, 2024), the role and emphasis on devolution in delivering better regional outcomes has never been greater. In the last 15 years, England has introduced over 10 combined authorities and metro mayors to streamline economic decision making so as to rejig the central-local framework that has existed for several decades. This paper attempts to shed some insights on the effectiveness of English devolution as currently constructed. The analysis presented should be seen as an early cross check of government devolution policy seeking to answer one fundamental question: has devolution and/or more devolved governance mechanisms led to better decision making, including and more specifically, better investment returns via better economic prioritisation of projects and investment initiatives. Crucially, the findings presented in the paper suggest no strong correlation between the level of devolution achieved by regions and returns to investment across England - be it in total gross fixed capital formation, ICT investment, intangible investment, 'other tangibles', and transport investment. Several robustness checks to cross check the baseline results were employed - all yielding similar results. Put simply, there is no positive and statistically significant link when it comes to devolution governance mechanisms in England and investment returns. The results uncovered adds to the empirical literature on English devolution in two ways. First, it is the only paper to our knowledge to codify regional governance mechanisms throughout the country. Second, it is also the first to make use of experimental regional investment statistics published by the ONS. In incorporating both elements, the paper adds to a scarce amount of quantitative research analysing the effects of devolution mechanisms and its impact on decision making and economic performance in the UK - something that has not been quantitatively investigated in much depth yet. To be sure, this is not to say that devolution has failed, however, or yielded sub-optimal outcomes. The quantitative analysis presents many challenges and caveats, which should be carefully considered. For instance, investment (GFCF) data are still classified as 'experimental' data. As with almost all national accounts statistics, data are prone to major revisions. Second, as the Institute for Government has highlighted, the effects of devolution may have yet to be fully felt, with some regions requiring years before we see any tangible differences in investment returns or indeed economic performance (Pope, et al., 2023). Specifically, given the time sample around which the analysis is based (2000-2019), any tangible policy effects may have yet to crystallise - particularly with infrastructure projects taking years for any benefits to be felt. Put simply, recent devolution policy (2010+) remains at its infancy and therefore may be harder to quantify at this juncture. A longer time-series of data may be needed to capture some of the devolution dividends associated with more effective economic prioritisation of projects and therefore growth. Moreover, changes in devolution are not random. Instead there is often political need for change – and the last decade or so has been mired with economic malaise meaning that change in governance mechanism may initially coincide with weak economic performance. Lastly, the model used to assess the link between devolution and investment returns may suffer from omitted variable analysis, namely in the lack of control for institutional capacity and quality of governance – which therefore raises further questions around central government strategy to push universally for devolution without tackling fundamental issues pertaining to the kind of devolution delivered across England. As things stand, however, the analysis adds credence to the existing literature that devolution in the UK (and in England in particular) has failed to capture any tangible and genuine benefits given the still elevated role of Treasury in determining funding and investment decisions. Changes in governance mechanisms as seen in the last decade do not necessarily entail policy control. The Institute for Government notes that despite devolution being “a necessary precondition for stronger local institutions”, there are costs to devolution, including loss of economies of scale (resulting in an increase in administrative costs) and harmful competition (particularly with regards to the prevailing deal making funding framework employed centrally), see Pope, et al., 2023. Effective policy coordination depends on thorough analysis and strong stakeholder collaboration, necessitating skilled professionals. However, due to England’s centralized governance and the traditional clustering of civil servants in London, some regions face shortages of qualified policy experts (see Centre for Cities, 2024, Guerin, et al., 2021). Local/regional governments also lack sufficient spending power to deliver beyond their core functions, limiting the effectiveness of devolution mechanisms created in the last decade or two. As currently constructed, the two primary tax levers for English local government are council tax and business rates – both of which comprise only around 25% of funding (well below the 40% average across other advanced economies (Haylen, 2019). Put simply, changes in the classification of sub-national government tiers do not imply effective change in policy making. The current funding for local/regional governments is essentially centred around competitively awarded grants. Local tax revenues play a very small role in funding day-to-day activities as well as investment. McCann (2022) highlights that the UK’s cost-based approach to funding local governments provides “much weaker fiscal stabilisation underpinnings than do revenue-based systems” with levels of subcentral government autonomy limiting local policy-making discretion while the UK’s over-centralised UK governance system “militates against both central government learning and local government institutional capacity building”. The Industrial Strategy Council too notes that subnational funding presented challenges given its competitive model with “stringent rules on spending along with ring-fenced budgets” limiting targeted, long-term interventions that would go a long way to address place-based issues while enabling longer-term strategic planning and implementation (Brittain & Taylor, 2021). McCann (2016) also concludes that “wholesale reform of the UK sub-central fiscal system must be a key part of an overall overhaul of the overly top-down and centralised UK governance system” with the “current UK central-subcentral fiscal system [representing] the worst of all worlds, in that it combines excessive centralisation with severe fragmentation”. In a recent report, the Resolution Foundation in partnership with the Centre for Cities (Breach & Bridgett, 2022) also found that “despite recent waves of devolution, Britain has become even more fiscally centralised since 2015, further reducing the already-weak incentive for local authorities to grow their local economies”. Providing better flexibility and medium-term funding certainty could go a long way in improving funding capabilities to deliver better investment returns. Lack of institutional capacity may have also hampered the abilities of combined (mayoral) authorities to reap any of the benefits of devolution. A dearth of local capacity could lead to uneven or less than optimal delivery. Pope et al. (2023) highlight the plight of mayoral authorities’ staffing inadequacies, with places like Greater Manchester employing over 2,000 people while Cambridgeshire and Peterborough employ just over 50. The same report also highlighted that even in advanced mayoralties like greater Manchester and West Midlands, there are fewer than one employee per 1,000 inhabitants (compared to Toronto and Frankfurt who have more than 15 employees per 1,000 inhabitants). The IfG conclude that “England’s current local institutions are not suited to maximise the benefits of devolution”. Coyle and Muhtar (2021) also argue that the lack of cross government coordination often impedes the effectiveness of decentralisation/devolution, with asymmetric knowledge flows passing from the bottom to the top, as opposed to ‘top-down’. The way that devolution is currently structured also raises important questions in delivering better outcomes. Fragmentation horizontally and vertically have led to less effective governance mechanisms. Others have noted explicitly that fiscal devolution is made harder by the fact that local authorities are inconsistent with local economies (see Breach & Bridgett, 2022, McCann, 2022). Local services are managed at various levels of political decentralisation, further complicating the UK’s devolution progress. When considering the current state of combined authorities, the Productivity Institute (Shaw, 2024) notes that the “smaller scale of combined authorities is also likely to create institutions with insufficient capacity and capability, as well as ‘weaker voices’ that are more easily ignored by government.” The lack of effective sub-national governance means that locally specific 'knowledge inputs' fail to reach government decision-makers (Coyle & Muhtar, 2021). Put simply, governance mechanisms – as currently structured – do little to improve regional leadership whilst also reducing local and regional accountability. Several conclusions can be taken from the above analysis. While devolution in theory entails many potential benefits, it is also apparent that its effects have not been dramatic in improving investment decision making across England. A shift to flexible funding arrangements may better able combined (mayoral) authorities to deliver on local objectives and growth. A shift away from away from competitive funding pots may also prove more effective in delivering longer-lasting economic transformation – one that takes time and one that requires long-term regional strategic vision. The ability to hire and train staff to build institutional capacity to improve governance quality may also be a necessary pre-condition for better economic returns from devolution. Increased devolution structures may also need to be better structured within the central-local establishment, whereby Treasury enables local government to have more autonomy in setting regional economic strategies, whilst also working with other authorities to create a more holistic meso-regional development policy agenda – one that effectively works to narrow the North-South divide and improve regional productivity rather than merely creating independent economic spatial silos. This paper should serve as a reminder for policy makers that devolution in and of itself is not a panacea to deliver better economic performance and improve regional outcomes. The lack of independent evaluation of present spatial policies, including and perhaps most importantly, devolution, raises concerns around past and present governments’ rush to devolve – with many seeing devolution as a ‘quick fix’ to treat all regional economic ailments. Getting existing devolution right should be the higher priority rather than rushing towards a broad-scale devolution policy that proves piecemeal in nature and ultimate is less effective – potentially even exacerbating spatial inequality in England. Declarations Author Contribution SR wrote the main manuscript text, figures, and implemented all the analyses. Acknowledgement This paper has greatly benefitted from comments and suggestions by Dr Johan Larsson and Professor Diane Coyle. All errors are my own. Data Availability I have used publicly available data to generate a time series of the English political devolution variable. Funding Declaration: No funding was received for this research. References Burret , H. T., Feld, L. P. & Schaltegger, C. A., 2022. Fiscal federalism and economic performance new evidence from Switzerland. European Journal of Political Economy, 74(102159). Acemoglu, , D. & Dell, M., 2010. Productivity Differences between and within Countries. American Economic Journal: Macroeconomics, Volume 2, pp. 169-188. Acemoglu, D., 2002. Technical Change, Inequality, and the Labor Market. Journal of Economic Literature, 40(1), pp. 7-72. Aschauer, D. A., 1989. Is public expenditure productive?. Journal of Monetary Economics, 23(2), pp. 177-200. Autor, , D., Levy, F. & Murnane, R. J., 2003. The Skill Content of Recent Technological Change: An Empirical Exploration. Quarterly Journal of Economics, 118(4), pp. 1279-1333. Banister, D. & Berechman, J., 2003. Transport Investment and Economic Development. 1st ed. London: Routledge. Bardhan, P. & Mookherjee, D., 2006. Decentralization, Corruption and Government Accountability. In: S. Rose-Ackerman, ed. International Handbook on the Economics of Corruption. s.l.:Edward Elgar Publishing. Benhabib, . J. & Spiegel, M. M., 1994. The Role of Human Capital in Economic Development: Evidence from Aggregate Cross-Country Data.. Journal of Monetary Economics, 34(2), pp. 143-173. Beveridge, S. W., 1942. Social Insurance and Allied Services. London: His Majesty's Stationary Office. Blöchliger, H. & Égert, . B., 2013. Decentralisation and Economic Growth - Part 2: The Impact on Economic Activity, Productivity and Investment, Paris: OECD Publishing. Bloom, N. et al., 2013. Does Management Matter? Evidence from India. The Quarterly Journal of Economics, 128(1), pp. 1-51. Breach, A. & Bridgett, S., 2022. Centralisation Nation, London: Resolution Foundation. Brittain, B. & Taylor, A., 2021. Devolution and Governance Structures in the UK, s.l.: West Midlands Regional Economic Development Institute. Britteon, P. et al., 2024. The impact of devolution on local health systems: Evidence from Greater Manchester, England. Social Science & Medicine, 348(116801). Bun, M. J. G. & Windmeijer, F., 2010. The weak instrument problem of the system GMM estimator in dynamic panel data models. The Econometrics Journal, 13(1), pp. 95-126. Calderon, C. & Serven, L., 2004. The Effects of Infrastructure Development on Growth and Income, Washington, D.C.: World Bank Policy Research Working Paper. Centre for Cities, 2024. Cities Outlook 2024, London: Centre for Cities. Ciccone, A. & Hall, R. E., 1996. Productivity and the Density of Economic Activity. American Economic Review, 86(1), pp. 54-70. Coyle , D. & Muhtar, A., 2021. UK’s Industrial Policy: Learning from the past?, Cambridge: University of Cambridge - Bennett Institute for Public Policy. Coyle, D. & Hampton, L., 2023. Twenty-first century progress in computing, Cambridge: Bennett Institute for Public Policy, University of Cambridge. Crain, M. & Oakley, L. K., 1995. The Politics of Infrastructure. Journal of Law and Economics, Volume 38, pp. 1-17. Crescenzi, R., Di Cataldo, M. & Rodrıguez-Pose, A., 2016. Government Quality and the Economic Returns of Transport Infrastructure Investment in European Regions. Journal of Regional Science, 56(4), pp. 555-582. Crescenzi, R. & Rodrıguez-Pose, A., 2012. Infrastructure and Regional Growth in the European Union. Papers in Regional Science, Volume 91, pp. 487-513. Easterly, W. R., 2002. The Elusive Quest for Growth: Economists' Adventures and Misadventures in the Tropics. 2nd ed. Cambridge: MIT Press. Enikolopov, R. & Zhuravskaya, E., 2007. Decentralization and political institutions. Journal of Public Economics, 91(11-12), pp. 2261-2290. Faguet , J.-P. & Pöschl , C., 2015. Is decentralization good for development? Perspectives from academics and policy makers. In: Oxford: Oxford University Press, pp. 1-29. Faguet, J.-P., 2021. Understanding decentralization: theory, evidence and method, with a focus on least-developed countries., London: LSE Working Paper. Fisman, R. & Gatti, R., 2002. Decentralization and corruption: evidence across countries. Journal of Public Economics, 83(3), pp. 325-345. Gennaioli, N., La Porta, R. & Lopez-de-Silanes, F., 2013. Human Capital and Regional Development.. Quarterly Journal of Economics, 128(1), pp. 105-164. GLA Economics, 2022. Devolution and economic growth in London: A rapid evidence review. , London: Greater London Authority. Glaeser, E. & Resseger, M. G., 2010. The Complementarity Between Cities and Skills.. Journal of Regional Science, 50(1), pp. 221-244. Guerin, B., Thomas, A., Clyne, R. & Vira, S., 2021. Finding the right skills for the civil service, London: Institute for Government. Hanushek, E. A., , E. A. & Kimko, D., 2000. Schooling, Labor-Force Quality, and the Growth of Nations.. American Economic Review, 90(5), pp. 1184-1204. Haylen, A., 2019. Local road maintenance in England, London: House of Commons Library. Henisz,, W. J., 2002. The Institutional Environment for Infrastructure Investment. Industrial and Corporate Change, Volume 11, pp. 355-389. Hirshman, O., 1970. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. Cambridge: Harvard University Press. Hitchcock, A., Borrows, M. & Harwich, E., 2017. Vive la devolution: devolved public-services commissioning, London: Reform. Laverty, K. J., 1996. Economic "Short-Termism": The Debate, the Unresolved Issues, and the Implications for Management Practice and Research. The Academy of Management Review, 21(3), pp. 825-860. McCann, P., 2022. The Fiscal Implications of 'Levelling Up' and UK Governance Devolution, London: NIESR Occasional Paper LXIII. Meloche, J.-P., Vaillancourt, F. & Yilmaz, S., 2004. Decentralization or Fiscal Autonomy? What Does Really Matter? Effects on Growth and Public Sector Size in European Transition Countries., Washington D.C.: World Bank. Ministry of Housing, Communities, & Local Government, 2024. English Devolution White Paper, London: UK Government. Muringani, J., Fitjar, R. D. & Rodríguez-Pose, A., 2019. Decentralisation, quality of government and economic growth in the regions of the EU, London: LSE Research Online. Oates, W. E., 1973. Fiscal Federalism. Public Choice, Volume 14, pp. 155-157. OECD, 2019. Making Decentralisation Work: A Handbook for Policy-Makers, Paris: OECD Publishing. Pope, T., Dalton, G. & Coggins, M., 2023. How can devolution deliver regional growth in England?, London: Institute for Government. Raja, S. I. & Larsson, J. P., 2024. Have City Deals delivered higher productivity in England? An empirical assessment of a broad-spectrum local growth policy.. Urban Studies. Redding, S. J. & Turner, M. A., 2015. Transportation Costs and the Spatial Organization of Economic Activity. Handbook of Regional and Urban Economics , Volume 5, pp. 1339-1398. Rodríguez-Pose, A. & Ezcurra, R., 2010. Does decentralization matter for regional disparities? A cross-country analysis. Journal of Economic Geography, 10(5), pp. 619-644. Rodríguez-Pose, A. & Ezcurra, R., 2011. Is fiscal decentralization harmful for economic growth? Evidence from the OECD countries. Journal of Economic Geography, 11(4), pp. 619-643. Rodríguez‐Pose, A. & Gill, N., 2005. On the ‘economic dividend’ of devolution. Regional Studies, 39(4), pp. 405-420. Shaw, J., 2024. Devolution: the importance of scale and coterminosity, London: The Productivity Institute. Sow, M. & Razafimahefa, I. F., 2015. Fiscal Decentralization and the Efficiency of Public Service Delivery, Washington D.C.: International Monetary Fund. Tiebout, C. M., 1956. A Pure Theory of Local Expenditures. Journal of Political Economy, 64(5), pp. 416-424. Tomaney, J., 2016. Limits of Devolution: Localism, Economics and Post-democracy. The Political Quarterly, Volume 87. Torrance, D., 2019. Introduction to devolution in the UK, London: House of Commons Library. Treisman, D., 2007. The Architecture of Government: Rethinking Political Decentralization. Cambridge: Cambridge University Press. UK Government, 2015. 2010 to 2015 government policy: UK economic growth, London: House of Commons. Zhao, C., Feng, F., Chen, Y. & Li, X., 2021. Local government competition and regional innovation efficiency: From the perspective of China-style fiscal federalism. Science and Public Policy, 48(4), pp. 488-489. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers invited by journal 24 Apr, 2026 Editor assigned by journal 23 Feb, 2026 Submission checks completed at journal 23 Feb, 2026 First submitted to journal 16 Feb, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8896200","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":628967169,"identity":"05a5ad5e-ce40-4f84-8c0d-1ec77ccf186f","order_by":0,"name":"Sanjay Raja","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvElEQVRIiWNgGAWjYDACZgbGxz8q4NwEorQwGzOcIUkLAwObNGMbKVrk23kfSBfOs5PXbWB++IGxLY2wFoPD7AbGM7clG247wGYswdiWQ4QWZjaGBN5tBxi3HWAwY2BsqyCsRb6ZjeEA75wD9tsOsH8jTgvDYTbGZt6GA4nbDvCAbCHGYYfZmBlnHEtO3naYp1gi4RwR3pfvP8b+40ONne224+0bP3woSybCYXDAzEBsRI6CUTAKRsEoIAgA7ZEzh3zayOwAAAAASUVORK5CYII=","orcid":"","institution":"University of Cambridge","correspondingAuthor":true,"prefix":"","firstName":"Sanjay","middleName":"","lastName":"Raja","suffix":""}],"badges":[],"createdAt":"2026-02-16 21:08:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8896200/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8896200/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108540493,"identity":"ac31803b-bf5a-418d-9956-baa8dc15611a","added_by":"auto","created_at":"2026-05-05 18:33:36","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":212831,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDevolution Scores\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSource: Author’s Calculations, Office for National Statistics\u003c/em\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8896200/v1/f8857ffb887f98e7ee2f7290.png"},{"id":108540494,"identity":"ca55ada9-2de8-4524-a377-98ae9d349129","added_by":"auto","created_at":"2026-05-05 18:33:36","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":38695,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDevolution Scores versus GFCF Growth\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8896200/v1/a0df8270c0b74a11cd37e0b1.png"},{"id":108804324,"identity":"cbfb40df-017e-451d-979c-0c4cd4669aab","added_by":"auto","created_at":"2026-05-08 15:19:20","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":808992,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8896200/v1/f4eaeb60-1636-4c20-b5e0-490b61c6ac47.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Has Political Devolution Led to Better Investment Returns? A Look at English Governance Mechanisms Over The Last Two Decades","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eDevolution has emerged as a cornerstone of UK regional economic policy over the past decade, reshaping governance and offering a pathway to address spatial inequalities through empowered local decision-making (Warner, et al., 2024). But has devolution in the past two decades delivered better economic returns? From the City Deals to the Growth Deals, the last decade has seen an increase in decentralisation, with national governments seemingly shifting away from top-down economic development to bottom-up place-based policies (O\u0026apos;Brien \u0026amp; Pike, 2020). As part of England\u0026apos;s devolution drive, local governments have taken back more control in transport, skills development, and housing, with city regions also able to control the economic prioritisation of investment projects. As of 2025, there were a total of eleven metro mayors in charge of combined authorities \u0026ndash; a eleven-fold increase since 2000.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eLittle has been done quantitatively to assess the impact of devolution in England on economic returns. With devolution seen as a solution for regional inequality, breaking the stronghold of centralised decision making, and allowing local leaders to tailor policies and investments to specific needs and opportunities, the need to unpack its effects (even at its early stages) has never been more important. Indeed, the empirical case for devolution enabling regions to act as policy laboratories, experimenting with new approaches and learning from each other to drive higher returns on investment and innovation via better resource allocation deserves more attention. The case for devolution is compelling: empowered localities make bold decisions that central governments avoid (GLA Economics, 2022). \u0026nbsp;International evidence on devolution and its effects highlight some positives, with Burret et al. (2022) finding that competitive elements of decentralisation in Switzerland lead to improved economic performance, while Muringani et. al. (2019) show that decentralisation boosts regional GDP growth in the EU but only if paired with high-quality regional governance, as effective local institutions enable better public goods delivery and policy responsiveness.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor such a policy relevant subject, this paper presents one of the first quantitative analyses looking at the impact of devolution (i.e. governance mechanisms) and investment returns in England, using novel and experimental data on regional gross capital formation between 2000 and 2019. Why investment returns as a measure of economic success? Simply, because devolution (i.e. the transfer of powers from central government to local government), in theory, should lead to better investment returns, both public and private, by enabling more responsive, place-based decision-making and reducing bureaucratic barriers for investment and growth. This can be achieved via better local knowledge and strategic alignment, integrated planning and infrastructure settlements, enhanced coordination, and most importantly local ownership of economic development. The paper looks to add to the empirical literature focusing on the returns to devolution, particularly at a time when focus on regional policy has increased significantly.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eI find no relationship between devolution and investment returns across the full range of categories. Put simply, I find no evidence that devolution has led to better decision making when it comes to the economic prioritisation of projects \u0026ndash; something that sits at odds with the UK\u0026rsquo;s aim to deliver better economic prioritisation of investment via devolution.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhat does this suggest? Unlike other studies looking at broader European trends highlighting positive investment returns from better governance (Crescenzi, et al., 2016) something about devolution in England seems to be not working \u0026ndash; at least as far as these mechanisms go. Indeed, in focusing on changes in regional governance mechanisms, least at its initial stages, we are yet to find any strong success stories coming from the recent devolution drive. And while there may be good reasons why more devolved regions have yet to show \u0026ndash; at least statistically \u0026ndash; any positive and significant returns, relative to less devolved or decentralised government mechanisms in place, our findings seem to chime with some of the long-standing issues prevalent when looking at English devolution. The lack of fiscal decentralization (i.e. revenue powers), institutional capacity, and poorly defined economic functional areas may indeed be limiting the potential positive effects traditionally seen via regional devolution. \u0026nbsp;As focus on UK productivity and the lingering productivity puzzle increases, more research will need to be undertaken to investigate what drivers may ultimately be needed to better devolution outcomes in relation to investment returns in England. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis paper is structured as follows. The following section sets out the context around devolution, investment and economic growth, whilst also briefly outlining the state of regional devolution in England, and how place-based policy focused on the prioritisation of investment projects in the last decade. The next section presents the empirical approach in estimating the economic returns of investment associated with the scale of devolution in England, alongside its limitations. Following this, the section concludes with descriptive analyses, baseline estimation results, and a series of robustness checks. \u0026nbsp;The last section presents the conclusions while discussing the \u0026nbsp;broader implications of the study.\u0026nbsp;\u003c/p\u003e"},{"header":"2.\tDevolution, Investment, Economic Growth","content":"\u003ch2\u003e\u003cstrong\u003ea. Devolution and Economic Growth\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe term \u0026lsquo;devolution\u0026rsquo; has often been reflected the decentralisation of resources from central governments to local governments. This, however, represents a narrow definition. As Rodriguez-Pose and Gill (2004) highlight, devolution \u0026ldquo;can assume various forms, ranging from decentralisation of power and legitimacy through to a mere delegation of responsibilities and financial duties\u0026rdquo;.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe link between political devolution and economic growth is complex with much of the focus on devolution based on fiscal decentralisation as opposed to political governance changes. Decentralization fosters policy experimentation and competition, as regions adopt best practices from peers (\u0026quot;laboratory federalism\u0026quot;). Tiebout (1956), for example, posited that decentralization fosters competition among regions, enhancing growth. The model argues that competition among local governments replaces political processes in achieving efficient public goods allocation. Individuals \u0026quot;vote with their feet\u0026quot; by relocating to jurisdictions offering their preferred mix of taxes and services, eliminating the need for centralized preference aggregation. Zhao et. al (2021) found that fiscal decentralization significantly promoted regional innovation efficiency in China, with both vertical competition (between central and local governments) and horizontal competition (among local governments) further enhancing regional innovation efficiency. The authors conlcuded that local governments with more fiscal autonomy can better target investments and policies to stimulate innovation, creating positive externalities for innovation participants.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRodriguez-Pose and Ezcurra (2010), however, showed that any positive effects coming fiscal decentralisation or competition were highly state-dependent. As part of their empirical analyses, they concluded that decentralization was not systematically associated with changes in regional disparities \u0026ndash; whereby in high-income developed countries, decentralization was generally linked to a \u003cem\u003ereduction\u003c/em\u003e in regional inequality, suggesting that devolved powers could help even out regional growth when institutional capacity and redistributive mechanisms were strong. In low and middle income countries fiscal decentralization was associated with a \u003cem\u003esignificant increase\u003c/em\u003e in regional disparities. Here, poorer regions often lacked the institutional capacity, tax base, and infrastructure to benefit from decentralization, leading to widening gaps.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eElsewhere, Meloche et. al. (2004) examined the importance of fiscal autonomy in evaluating the effects of decentralization. Using data from European transition countries they found that fiscal autonomy\u0026mdash;where subnational governments have control over their own revenues - was positively associated with economic growth. The study distinguished between decentralisation (the devolution of functions) and true fiscal autonomy, finding the latter to be more growth-enhancing. A separate study by the OECD (Bl\u0026ouml;chliger \u0026amp; \u0026Eacute;gert, 2013) analysed the relationship between fiscal decentralisation and economic activity concluding that fiscal decentralisation was positively and significantly related to GDP per capita, productivity, and human capital. They deemed that revenue-based \u0026nbsp;decentralisation (tax autonomy) yielded stronger positive effects than spending-based measures. However, the relationship was non-linear, with diminishing returns as decentralisation increased, while intergovernmental transfers while negatively associated with growth.\u003c/p\u003e\n\u003cp\u003eWhile shifting to political devolution mechanisms, however, the literature has centred around improved efficiency in public sector delivery as well as increased accountability/less corruptibility, allowing regions to better enact place-based policies to drive growth and foster more efficient investment outcomes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOates (1973) via the \u0026ldquo;decentralisation theorem\u0026rdquo; posited that local governments were better able to match public goods to heterogeneous preferences, supported by welfare gains in heterogeneous jurisdictions. Oates argued that subnational governments could better tailor policies to local preferences, improving efficiency. \u0026nbsp;Bardhan \u0026amp; Mookherjee (2006) concluded that decentralisation shifted accountability closer to citizens but risked capture where some locals could disproportionately influence service allocation. Hooghe et al. (2010) showed that regions with high levels of devolution (looking at institutional depth, fiscal autonomy, and law-making powers) were better able to align services like education and healthcare with local needs. A separate paper from the International Monetary Fund (IMF) highlighted that fiscal decentralization could lead to improvements in public sector delivery efficiency - but only under specific conditions, including where regions/localities had adequate political and institutional environments, a minimum degree of spending decentralization and a suitable degree of revenue decentralization (Sow \u0026amp; Razafimahefa, 2015).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eEvidence from Nordic countries, for example, show that municipalities with control of healthcare and education achieved better outcomes, including satisfaction (Hitchcock, et al., 2017). Faguet and P\u0026ouml;schl (2015) found that decentralisation in Bolivia improved public service alignment with local preferences, enhancing access and quality in education, healthcare, and infrastructure. In the UK specifically, Britteon et al. (2024) concluded that the Greater Manchester devolution deal lead to improved population health outcomes, driven by improvements in health services and wider social health determinants.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMoreover, local governments have better information about their respective region\u0026rsquo;s needs and priorities and ultimately can be held more accountable to its electorate. Voters can reward/punish local leaders based on service quality. Faguet (2021) noted that decentralisation brings government \u0026quot;closer to the people,\u0026quot; shortening and tightening the loop of accountability. Local officials become directly accountable to citizens, whose votes and voices become more powerful in influencing local service delivery and policy outcomes. Hirschman (1970) described this as an \u0026ldquo;exit and voice\u0026rdquo; mechanism, whereby citizens can monitor the quantity and quality of services, and discipline politicians by rewarding or sanctioning them in competitive elections. \u0026nbsp;The empirical research surrounding this question has broadly confirmed the relationship between decentralisation/devolution, governance quality, and enhanced accountability. For instance, Fisman and Gatti (2002) examined the effects of decentralization and corruption \u0026nbsp; via a cross-country analysis, concluding that \u0026ldquo;fiscal decentralization in government expenditure is strongly and significantly associated with lower corruption\u0026rdquo;. Bardhan and Mookherjee (2006) noted that decentralization shortened the chain between citizens (principals) and officials (agents), making it easier to monitor and sanction corrupt behaviour. The OECD also showed that decentralization can improve service delivery and accountability when paired with strong local institutions, transparency, and social capital (OECD, 2019). Others have also argued that federal systems with strong local oversight (e.g., Germany, Canada) exhibit lower corruption than highly centralized ones.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eb. Devolution in England \u0026ndash; a brief look back at the last two decades\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eDevolution and regional governance in England have a complex and fragmented history. The idea of English regional devolution dates back to the early 20th century, with Winston Churchill advocating for regional parliaments in 1912 (Torrance, 2019). \u0026nbsp;Post-war reconstruction, guided by the Beveridge Report (Beveridge, 1942) and Keynesian economics, saw national government directly planning housing, welfare, and economic development. The National Health Service Act 1946 transferred health services from local governments to a centrally run NHS under the Ministry of Health. After World War II, several commissions, notably the Redcliffe-Maud Report (1969), recommended replacing England\u0026rsquo;s two-tier local government with larger unitary authorities grouped into eight provinces with devolved powers. However, these proposals were not implemented, as political focus shifted elsewhere. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn 1994, the Major government established ten Government Office Regions to coordinate central government activities regionally.\u0026nbsp;The Blair government (1997\u0026ndash;2010) advanced regionalism by creating Regional Development Agencies (RDAs) in 1998 to foster economic growth, and indirectly elected Regional Assemblies or Chambers to oversee them.\u0026nbsp;However, these assemblies lacked direct democratic legitimacy. A key moment came in 2004, when a referendum in North East England overwhelmingly rejected a proposed elected regional assembly.\u0026nbsp;This defeat led to the abandonment of plans for further regional assemblies, and between 2008 and 2010, the existing assemblies and RDAs were abolished. The one major devolution success, however, came via the capital. The Greater London Authority (GLA), comprising a directly elected Mayor and Assembly, was established in 2000, marking the only successful city-regional devolution in England. Most councils until this point relied on a \u0026lsquo;two-tier government\u0026rsquo; as per a \u0026lsquo;leader-cabinet model\u0026rsquo; where councillors were elected and subsequently elected a leader.\u003c/p\u003e\n\u003cp\u003eSince 2010, the focus for national governments shifted to \u0026ldquo;City Deals\u0026rdquo;, \u0026ldquo;Growth Deals\u0026rdquo;, and more-recently, the \u0026ldquo;Levelling-Up\u0026rdquo; agenda. The push towards devolution took the form of combined authorities and directly elected metro mayors. \u0026nbsp;Combined authorities enabled groups of local councils to voluntarily collaborate, often led by directly elected mayors (e.g., Greater Manchester, West Midlands). These combined authorities gained devolved powers over transport, housing, and economic development. Some combined authorities even received increased fiscal decentralisation via a full scale (pilot) 100% business rates retention scheme (including Greater Manchester, Liverpool, West of England and the West Midlands). Devolution deals towards the end of the decade also included commitments to multi-year funding for specific projects, though recent waves of devolution have invoked some criticism. The Industrial Strategy Council recently noted that policies regarding sub-national governance \u0026ldquo;have been piecemeal rather than coherent strategies\u0026rdquo;, resulting in a \u0026ldquo;current local institutional landscape [that] is confusing and fragmented\u0026rdquo;, having \u0026ldquo;no clear, long-term, coherent vision for sub-national policy\u0026rdquo;.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003ec. English devolution and economic returns of investment\u0026nbsp;\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eEconomic policies need to be tailored to their local needs, given their respective industrial bases, demographics, human capital, and local expertise. In theory, local governments will almost always be better able to identify and prioritise economic projects \u0026ndash; especially in the context of constrained and limited funding. This was the goal for better and deeper political devolution in England over the last several decades \u0026ndash; and in particular, over the course of the last two decades where governments (past and present) sought to push local intervention in the prioritisation of economic projects in a bid to ramp up regional productivity convergence and growth.\u003c/p\u003e\n\u003cp\u003eThe proliferation of so-called \u0026lsquo;competitive bids\u0026rsquo; and \u0026lsquo;spending pots\u0026rsquo; required local governments (either via Local Enterprise Partnerships, City Deals, Growth Deals, and or Trailblazer Deals) to bid or make applications to central government to receive grant funding \u0026ndash; where funding would be determined on the quality of potential economic returns associated with local projects and initiatives. \u0026nbsp;Between 2010 and 2024, economic prioritisation of investment projects have come via a number of place-based policy schemes in an effort to maximise impact and improve economic returns (UK Government, 2015) via Local Enterprise Partnerships (LEPs), City and Growth Deals, and Enterprise Zones. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe importance of political decentralisation combined with local authorities will almost always be better able to channel local expertise and knowledge as well as fostering the \u0026lsquo;right\u0026rsquo; type of investment \u0026ndash; and most importantly, investment projects that align with each region\u0026rsquo;s local strategic. The capability to better identify and prioritise high-return projects, target investment into sectors with regional comparative advantage, and respond more nimbly to local needs can boost investment returns in not just the public sector, but the private sector too. Furthermore, the ability to strike multi-year funding arrangements can also help \u0026lsquo;crowd-in\u0026rsquo; private sector investment in contribution to various regions\u0026rsquo; local ambitions. The introduction of long-term planning can reduce investment risk more broadly across regions for both public and private actors. Integrated strategies for transport, housing, and skills are critical for unlocking private sector investment and improving public investment efficiency. The fostering of strong partnerships between public bodies and private investors could enable co-investment and leveraging public funds to crowd in private capital, making regions more attractive to investors.\u0026nbsp;\u003c/p\u003e"},{"header":"3.\tEmpirical Model and Regression Estimations","content":"\u003ch2\u003e\u003cstrong\u003ea. Model Specification and Data\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe empirical analysis aims to shed any insight on whether political devolution \u0026nbsp;helped shape investment returns across England given the implicit focus around the economic prioritisation of investment projects. The analysis relies on multiple channels of investment using novel and experimental investment data from the Office for National Statistics (ONS). Including various categories of investment (i.e. gross fixed capital formation) allows for an investigation that reflects structural differences in specific regional investment categories, allowing for a more thorough investigation of various investment decisions based on regional/local political preferences, the financial requirement to deliver them, and consequently, the potential impact on regional economic growth.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe impact of devolution is modelled by including an interaction term between proxies for investment and the scale of English devolution. The following model is used to estimate the link between devolution and investment returns on economic growth:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"https://myfiles.space/user_files/58895_8739fc6c57c1c19a/58895_custom_files/img1778005419.png\" width=\"825\" height=\"373\"\u003e\u003c/p\u003e\n\u003cp\u003eUsing novel and experimental data from the ONS, the paper utilizes the following breakdown of gross capital fixed formation as proxies for regional investment:\u0026nbsp;\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003e\u003cem\u003eBuildings and Structures\u003c/em\u003e: This includes everything from dwellings, roads and any other construction of buildings and structures;\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eTransport Equipment\u003c/em\u003e: This includes ships, road transport vehicles, and all other transport equipment (including aircrafts);\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eICT Equipment\u003c/em\u003e: This includes computer hardware and telecoms;\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e\u003cem\u003e\u0026lsquo;Other Tangible Assets\u0026rsquo;\u003c/em\u003e: This includes machinery and equipment, cultivated assets, and weapon systems;\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eAnd Intangible Assets\u003c/em\u003e: This includes research and development (R\u0026amp;D), mineral exploration and evaluation, computer software and databases (both purchased and own account software), as well as entertainment, literary or artistic originals.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eAll of the above investment proxies are expressed in the form of a natural logarithm and are normalised using regional population data (i.e. investment per capita). All data are deflated using relevant regional deflators. Turning to the political devolution proxy, and what is a new contribution to the literature, a \u0026nbsp;political devolution variable is constructed based on a numerical score of 1-5 between 2004-2019 with 1 representing the lowest form of political devolution and 5 the highest form of political devolution. The scores assigned are attributed as follows:\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e1 = Two-Tier Local Council \u0026ndash; the lowest form of devolution. The Two-Tier Local Council has decision making shared between the two tiers of local government (upper tier being county councils, and lower tier being district and borough councils).\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e2 = Cabinet System. This is a small group of elected councillors, where power is limited but concentrated in the cabinet.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e3 = Unitary Authority. This is where a single council handles all local government functions (replacing the two-tier system).\u003c/li\u003e\n \u003cli\u003e4 = Combined Authority. This includes multiple local authorities to oversee regional issues, with transfer of certain powers from central government to the combined authority enabling regional decision-making.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e5 = Elected Mayor \u0026ndash; the highest form of devolution. This represents a directly elected mayor that leads the local authority or combined authority, providing leadership and accountability for local and regional governance.\u0026nbsp;\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThe choice of political devolution correspond to meaningful structural reforms in governance, with each level representing incremental increases in both legal and decision-making powers transferred from central government to local government (see \u0026nbsp;(Local Government Association, 2024)). Two-tier local councils and cabinet-style systems, for instance, have limited control over basic services, planning and regulatory functions. Unitary authorities and combined authorities have greater control over major public services, economic development, and infrastructure. While elected metro-mayors provide executive leadership and accountability with increased discretion over locally-raised and devolved funds, policy priorities, and regional coordination. Combined authorities as well as elected metro mayors see additional fiscal autonomy too, with powers over transport, skills, housing, adult education, as well as \u0026ndash; in some cases \u0026ndash; increased ability around tax retention.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo understand how devolution affects investment, an interaction of the devolution variable and investment is used. The spatial level at which the model specification is run on is the International Territorial Level 3 (ITL3) level yielding a more granular - yet economically relevant - level of spatial analysis on devolution and economic returns.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe vector of controls includes a number of variables traditionally linked to impacting economic growth. In accordance with the endogenous growth theory, we include proxies for sectoral composition, agglomeration effects, human capital, jobs density, spillovers of regional investment. The sectoral composition is defined as the GVA share for production, services (excluding public services), and public services. Agglomeration effects are proxied for by population density \u0026ndash; defined formally as population per square kilometre (see Ciccone \u0026amp; Hall, 1996, Glaeser \u0026amp; Resseger, 2010). Human capital is proxied by the share of the regional population with NVQ4+ education (Benhabib \u0026amp; Spiegel, 1994, Hanushek, E. A., \u0026amp; Kimko, 2000, Gennaioli, et al., 2013). Jobs density acts as a proxy for labour demand \u0026ndash; another important determinant in driving economic growth (Acemoglu, 2002, Autor, , et al., 2003), and is defined as the number of jobs per working age population (those aged 16-64). Given that investment in buildings/structures, transport, ICT, tangible assets and intangible assets will likely impact economic growth and performance beyond the specific area in which the investment takes place, we also account for any spillovers generated from investment with a spatial lag. Estimation of the spatial lag is based on an exponential decay function, similar to the one used by Raja and Larsson (2024). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo deal with the changing nature of geographical data in the UK, a uniquely constructed ITL3 dataset is used, comprising of 114 regional areas. For all 114 ITL3 areas, there is a fully balanced dataset. The time period used in the baseline estimations is 2000-2019 \u0026ndash; providing nearly two decades of data.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eb. Descriptive Statistics\u0026nbsp;\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eTable 1 presents a summary of the key statistics included in the dataset, highlighting both the mean and standard deviation of key variables used in the econometric specification in years 2000 and 2019 for comparison. Looking across all 114 ITL3 regions constructed for the analysis, average regional GVA grew by roughly 2% per annum between 2000 and 2019. The broad sectoral composition of the economy also shifted as one would expect, with the share of production across all ITL3 regions narrowing by 2 percentage points. The services share across all ITL3 regions increased by 5 percentage points. And the public sector share of all ITL3 regions dropped by 2 percentage points.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eLooking at investment levels, one can see similar increases in investment levels (GFCF), with total gross fixed capital formation levels growing by roughly 2% per annum across the time sample. Looking at the disaggregated breakdown of investment, buildings growth saw the largest growth at 3.8% per annum. Intangibles investment growth averaged nearly 3% per annum. \u0026lsquo;Other tangibles\u0026rsquo; and intangibles investment barely grew. And ICT investment contracted between 2000-2019, reflecting potential falls in cost over time (see Coyle \u0026amp; Hampton, 2023) (Coyle \u0026amp; Hampton, 2023).\u003c/p\u003e\n\u003cp\u003eOther included variables track generally as expected over the last two decades. Jobs density (i.e. the number of jobs per working age population), for example, increased by 0.07pts. The share of the population with NVQ 4+ qualifications increased by 17 percentage points over the same time period. And the population density across ITL3 regions increased by nearly 400 people per square kilometre.\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"718\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 1: Descriptive Statistics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" colspan=\"3\" valign=\"bottom\"\u003e\n \u003cp\u003e2000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" colspan=\"3\" valign=\"bottom\"\u003e\n \u003cp\u003e2019\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eObservations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eStd. Deviation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eObservations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eStd. Deviation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGVA (\u0026pound;, millions, real)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e10,690\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e8,629\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e15,056\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e14,502\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGVA, Production Share\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e17%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e8%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e15%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e8%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGVA, Services Share (excl public servcies)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e53%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e11%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e58%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e10%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGVA, Public Services Share\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e23%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e7%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e21%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e6%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eJobs Density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e0.49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e0.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eSkills Share (NVQ 4+)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e23%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e7%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e39%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e11%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003ePopulation Density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e2,205\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e2,521\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e2,601\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e3,116\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGFCF Total Investment (\u0026pound;, millions, real)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e2,133\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e1,597\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e3,054\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e2,590\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGFCF Buildings Investment (\u0026pound;, millions, real)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e1,007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e661\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e1,734\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e1,394\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGFCF Intangibles Investment (\u0026pound;, millions, real)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e441\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e493\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e685\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e913\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGFCF Other Tangibles Investment (\u0026pound;, millions, real)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e337\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e290\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e374\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e371\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGFCF ICT Investment (\u0026pound;, millions, real)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e209\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e179\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e116\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e123\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003eGFCF Transport Investment (\u0026pound;, millions, real)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e140\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e103\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e145\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e147\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cem\u003eSource: Office for National Statistics, NOMIS\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTurning to the devolution variable, Table 2 presents a summary table of the five basic categories of political and economic devolution used in the analysis to depict devolution governance mechanisms across England: two-tier local councils, cabinet system, unitary authorities, combined authorities, and elected mayors. Since 2010, there has been a marked increase in devolution with several ITL3 regions falling into either a combined authority or metro-elected mayor governance mechanism. In 2019, 56 regions are identified as part of either a combined authority or part of a metro-elected mayoralty \u0026ndash; three times as many as in 2010.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFigure 1 presents a spatial visual of the assigned devolution scores across England. These are based on the final devolution mechanisms in place for ITL3 regions as of 2019 \u0026ndash; the final year of the analysis. Figure 2 presents a scatter-plot of regional devolution scores at the ITL3 level, relative to the cumulative investment (GFCF) growth between 2000-2019. A visual inspection of the data suggests no obvious relationship between devolution level and investment growth but we do see those ITL3 regions at the top end of the devolution spectrum registering the largest cumulative increase in investment growth. \u0026nbsp;\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"468\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" colspan=\"4\" valign=\"bottom\" style=\"width: 468px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 2: Governance Mechanism By Year\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 337px;\"\u003e\n \u003cp\u003eGovernance Mechanism, # of ITL3 Regions\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e2010\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e2019\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e2024\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 337px;\"\u003e\n \u003cp\u003eTwo-Tier Local Council\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 337px;\"\u003e\n \u003cp\u003eCabinet System\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 337px;\"\u003e\n \u003cp\u003eUnitary Authority\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e37\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 337px;\"\u003e\n \u003cp\u003eCombined Authority\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 337px;\"\u003e\n \u003cp\u003eMetro Elected Mayor\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 44px;\"\u003e\n \u003cp\u003e54\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cem\u003eSource: Author\u0026rsquo;s calculations\u003c/em\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003ch2\u003ec. \u003cstrong\u003eEstimation Issues and Specifications\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe empirical model outlined in Equation (1) is estimated using fixed effects panel methods, incorporating time dummies to account for temporal variations. To address potential issues of serial correlation and heteroscedasticity, clustered standard errors are employed (at the ITL3 level). Additionally, the impact of spatial autocorrelation\u0026mdash;where error terms of neighbouring observations may not be independent\u0026mdash;is mitigated by including spatially lagged variables as controls. These variables explicitly capture interactions between neighbouring regions, thereby reducing their influence on the residuals.\u003c/p\u003e\n\u003cp\u003eUsing\u0026nbsp;fixed effects\u0026nbsp;in a panel regression controls for unobserved and time-invariant heterogeneity across all ITL3 regions, ensuring that the estimated coefficients are not biased by omitted variables that differ across entities but remain constant over time. Including\u0026nbsp;time\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eeffects\u0026nbsp;accounts for common temporal shocks or trends that affect all entities simultaneously, isolating the impact of variables of interest from broader time-specific influences. Together, fixed effects and time effects enhance the robustness and accuracy of the regression results by addressing both cross-sectional and temporal variations.\u003c/p\u003e\n\u003cp\u003eThe regression specification also tries to mitigate against omitted variable bias through the inclusion of temporal lags \u0026ndash; in this case, GVA per capita. Lagged GDP/GVA growth can act as a proxy for unobserved factors that affect current growth, such as institutional quality, cultural factors, or historical events. Moreover, given that GDP/GVA growth rates are often serially correlated, including lags helps account for this autocorrelation, improving the efficiency of the estimates (see Beck and Katz, 1995; Arellano \u0026amp; Bond, 1991).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe baseline estimations are provided in Table 3. Column 1 includes a stripped down version of the specification, including only the lagged GVA growth of a region and the devolution variable. Column 2 introduces total gross fixed capital formation investment growth. Columns 3-7 include disaggregated investment categories. In columns 2-7, the regression specification includes the devolution proxy, investment growth, and the interaction between the devolution variable and investment growth. The interaction term tests whether the\u0026nbsp;effect of investment growth on GDP growth differs\u0026nbsp;in regions with varying levels of devolution. Put differently, the specification tests whether investment returns are stronger as a result of increasing devolution.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003ed. Baseline Estimations\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eIn all specifications, the devolution variable is positive, but not a statistically significant factor when it comes to driving GVA per capita growth. Changes in investment (total, intangibles, buildings, other tangibles, ICT, transport) are all positively linked to GVA per capita growth \u0026ndash; but are not statistically significant. The key devolution variable interacted with changes in investment (be it total investment or disaggregated investment) is also not statistically significant in any of the investment specifications included\u0026ndash; and in some cases (i.e. total and transport) are negatively signed.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhat variables show up as statistically significant? Lagged GVA per capita \u0026ndash; a sign of persistence is negatively signed but always statistically significant. Spatial spillovers of investment (intangibles, other tangibles, ICT) have a statistically significant relationship with GVA per capita growth. Industry shares also show up as statistically significant.\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" align=\"\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" colspan=\"8\" valign=\"bottom\" style=\"width: 602px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 3: Baseline Regression Results\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eDependent Variable: Change of Log GVA Per Capita\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cem\u003eTotal\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cem\u003eIntangibles\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cem\u003eBuildings\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cem\u003eOther Tangibles\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cem\u003eICT\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cem\u003eTransport\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eLagged GVA Per Capita\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e-0.073***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.182***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.181***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.182***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.182***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.182***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.182***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e(0.0167)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0248)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0247)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0248)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0249)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0250)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0258)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eDevolution\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e-0.000515\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e9.37e-05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-5.00e-05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e6.28e-05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e5.69e-05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e6.45e-05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.000103\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e(0.00103)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.000946)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.000951)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.000951)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.000944)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.000945)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.000955)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eChange Investment (GFCF) Per Capita\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.00759\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.00213\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.000536\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.000270\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.00122\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.00473\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00527)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00450)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00416)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00304)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00241)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00410)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChange Investment Per Capita x Devolution\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.000425\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.00170\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.000324\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.000800\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.000278\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.000248\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e(0.00172)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e(0.00171)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e(0.00136)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e(0.000959)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e(0.000780)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e(0.000878)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eChange in Spatial Weight of Investment Per Capita\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.000112\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.00505**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.00206\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.00316**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.00285**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.00102\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00303)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00221)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00232)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00153)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00143)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.00144)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eProduction Share (GVA)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e1.88e-05***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e1.87e-05***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e1.90e-05***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e1.92e-05***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e1.91e-05***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e1.92e-05***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(3.54e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(3.52e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(3.53e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(3.55e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(3.55e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(3.66e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eServices Share (GVA)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e4.51e-06***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e4.52e-06***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e4.55e-06***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e4.48e-06***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e4.49e-06***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e4.35e-06***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(8.38e-07)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(8.32e-07)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(8.42e-07)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(8.52e-07)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(8.53e-07)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(9.21e-07)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003ePublic Sector Share (GVA)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e7.06e-06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e7.04e-06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e6.72e-06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e7.05e-06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e7.06e-06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e8.38e-06\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(5.88e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(5.92e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(5.92e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(5.86e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(5.87e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(6.07e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003ePopulation Density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-1.01e-05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-1.03e-05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-1.01e-05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-8.01e-06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-7.98e-06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-9.90e-06\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(7.14e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(7.02e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(7.13e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(6.67e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(6.70e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(7.58e-06)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eJobs Density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.0190\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.0200\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.0188\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.0182\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.0182\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.0174\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0187)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0187)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0188)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0190)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0190)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0189)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eHuman Capital\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.0278\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.0281\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.0274\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.0218\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.0216\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.0176\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0352)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0350)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0351)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0350)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0350)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.0374)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eConstant\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e-0.251***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.705***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.701***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.707***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.710***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.710***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.708***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e(0.0622)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.103)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.103)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.103)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.104)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.104)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e(0.109)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eObservations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e2,280\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e2,280\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e2,280\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e2,280\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e2,280\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e2,280\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e2,280\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eR-squared\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e0.613\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.471\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.473\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.470\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.472\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.472\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e0.480\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eYear FE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 85px;\"\u003e\n \u003cp\u003eNumber of ITL3 Regions\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e114\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eRobust standard errors in parentheses; \u0026nbsp;*** p\u0026lt;0.01, ** p\u0026lt;0.05, * p\u0026lt;0.1\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe insignificant relationship between the devolution variable, investment changes, and the interaction of devolution variable and investment (i.e. investment returns) all suggest that the link between the devolution variable and investment returns is less clear cut. There could be several reasons for this. For one, identification of such a relationship across 114 unique ITL3 regions may be complex and difficult to determine. As Cresenzi et al. (2016) noted, optimal investment would lead to zero marginal returns, given the potentially efficient allocation of resources, leading to a potentially weak statistical relationship between devolution and investment returns on paper. Diminishing marginal returns to capital may also play a role in the weaker statistical relationship found between the key variables of interest \u0026ndash; though given the UK\u0026rsquo;s productivity problems, this seems less likely. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhat are other possible conclusions? Easterly (2002) cite that poor institutions could lead to capital misallocation, reducing any returns to investment. This would suggest that the link between deeper devolution and governance quality/institutional capacity may be weak \u0026ndash; with regional institutions in the UK potentially lacking the necessary institutional structures and effective governance mechanisms to deliver better investment returns. Another potential explanation lies with the notion of poor management practices. Bloom et al (2013) show that poor management practices lower returns. Another potential explanation could also be with regards to the decision making processes within English regions. The introduction of deeper devolution to England across the country could have led to poor economic prioritisation, and in some cases leading to myopic investment decisions \u0026ndash; decisions that may highlight important economic importance, but in reality do little to actually lift economic growth and/or productivity, see Laverty (1996) and Rodr\u0026iacute;guez-Pose \u0026amp; Gill (2020). \u0026nbsp;Tomaney (2016) also noted that devolution in the UK did not show an equal shift in democratic advances with recent devolution programmes \u0026ldquo;embedded within a centrally imposed tax cutting agenda\u0026rdquo; lacking \u0026ldquo;fiscal solidarity and a genuine attempt to extend democratic accountability\u0026rdquo;. \u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003ee. Robustness Checks\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eIn this section, several robustness checks are deployed to assess whether the above findings hold. In particular, the robustness checks employed include: a re-definition of the devolution variable, changing of the empirical time specification, change in dynamic specification, using alternative proxies for investment, and dealing with potential endogeneity concerns. Each are taken in turn below with results presented in Tables A1-A5.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eRedefining the devolution variable\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e.\u003c/em\u003e One check to assess whether the above findings hold is by re-defining the devolution variable into a binary variable. Here only the two highest forms of devolution are assessed: combined authorities and elected mayors. In re-defining the devolution variable to only include the highest forms of devolution, measurement error and heterogeneity bias across governance mechanisms is minimised. Pooling all devolved regions together (including those with weak powers) could dilute the estimated \u0026lsquo;treatment effect\u0026rsquo;, leading to attenuation bias (measurement error) or spurious null results. Restricting the analysis, therefore, to ITL3 regions with meaningful devolution (including those combined authorities that saw increased fiscal decentralisation via 100% business retentions including West Midlands, Greater Manchester, and Liverpool) would ensure that the devolution variable defined in the empirical estimations is strong enough to detect any effect.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThere is empirical support to such a claim. Rodriquez-Pose and Ezcurra (2011) show large heterogeneity when looking at the effects of fiscal decentralisation and economic growth. Triesman (2007) showed that shallow devolution (e.g., administrative delegation) had no measurable economic impact, creating noise in pooled estimates. Furthermore, Enikolopov \u0026amp; Zhuravskaya (2007) showed that only strong devolution (with electoral accountability) improved governance and growth, while weak decentralization worsened economic outcomes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn re-estimating the baseline specification with the new devolution variable, the findings do not change \u0026ndash; see Table A.1. The devolution variable, which is now a binary variable, is positive, but remains statistically insignificant. The key interaction variable in the empirical estimation shows mixed signs depending on the type of investment specified but crucially also remains statistically insignificant \u0026ndash; reaffirming the findings in our baseline estimations.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eChanging the time specification.\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAnother robustness check employed to validate the baseline estimation findings is by re-defining the time specification used for the empirical analysis. Here, instead of using the full dataset (2000-2019), a shorter timeframe is used. Specifically, the baseline estimations are re-run using a shorter timeframe of 2010-2019 \u0026ndash; the period in which most devolution agreements in our dataset were agreed. This would also remove the period in which the global financial crisis occurred. Similar to the baseline estimations, both the devolution variable and the interaction between the devolution and change in respective investment variable remain statistically insignificant (see Table A.2). In fact, in all six specifications, the key devolution interaction variable with investment change is negative.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eChange in dynamic specification.\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eThe baseline model specified is a dynamic specification where the lagged level of GDP is used as a regressor. This, in effect, attempts to account for convergence as well \u0026nbsp;as controlling for initial economic/GDP conditions of a region. As a further robustness check, a parsimonious version of the model is used, whereby the lagged GVA of a region is excluded. The results are presented in Table A.3. Again, the key interaction variable between devolution and change in investment remains statistically insignificant. The devolution variable is negative regardless of investment type specified.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAlternative proxies for investment: road investment.\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eIn a similar analyses conducted by Crescenzi et al. (2016), the baseline equation is re-estimated using road investment data \u0026ndash; instead of regionally disaggregated investment data.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn England, the level of autonomy local and combined authorities have over transport projects depends on the specific devolution deal agreed with central government. Mayoral Combined Authorities, for instance, created under the Cities and Local Government Devolution Act 2016, often receive the widest set of powers. These can include control over bus services through franchising (as in Greater Manchester and the West Midlands), responsibility for developing and delivering a local transport plan, management of key road networks, and influence over local rail services in partnership with the Department for Transport. They may also receive devolved funding streams\u0026mdash;such as the Transforming Cities Fund or City Region Sustainable Transport Settlements\u0026mdash;allowing them to set long-term investment priorities without seeking central government approval for every project. Non-mayoral combined authorities and individual local authorities generally have narrower transport powers, often focused on local road maintenance, parking, cycling and walking schemes, and some bus service support, but they typically lack full control over public transport integration or multi-year transport budgets.\u003c/p\u003e\n\u003cp\u003eAs a proxy for infrastructure investment, the re-estimated model focuses on the following transport investments at the ITL3 level: total road length, motorways, A-roads, minor roads, trunk roads, and principle roads. Why look at transport investment in particular? Good transport infrastructure enhances local accessibility, and leads to increased activity via a reduction in transport costs and increased productivity (see Aschauer, 1989, Banister \u0026amp; Berechman, 2003, and Redding \u0026amp; Turner, 2015). As Crain and Oakley, 1995; Henisz, 2002; Acemoglu and Dell, 2010 highlight, local institutions \u0026ndash; and by extension, regional governance mechanisms \u0026ndash; that drive the state of incentives and constraints are effectively moulded by local institutions and the political capital of regions to maximise the returns to transport investment (Crain \u0026amp; Oakley, 1995, Henisz,, 2002, Acemoglu, \u0026amp; Dell, 2010). As Crescenzi et al. (2016) argues, \u0026ldquo;[inadequate] political institutions may negatively affect the economic returns to transport infrastructure investment well before the money is actually spent. Governments are directly responsible for appropriate infrastructure planning and rigorous project selection, making transport infrastructure planning and financing fundamentally a political topic.\u0026rdquo;\u003c/p\u003e\n\u003cp\u003eAs such, in using transport road data as proxies for transport investment, this particular robustness check assesses whether the earlier findings hold for a different proxy of investment. Results of the regression results are outlined in Table A4. Consistent with above findings, no statistically significant relationship is uncovered when looking at both the devolution variable and the interaction variable between devolution and change in transport investment. What does this suggest? Even when altering the investment variable, the model specification shows no statistically significant correlation between devolution and investment returns associated with transport/road investment, implying very little impact from English devolution on the economic prioritisation of transport investment \u0026ndash; consistent with the baseline results.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eDealing with endogeneity: a system GMM model.\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eThe estimated effect of devolution and changes in investment (proxied by various investment categories within gross fixed capital formation) on economic growth may be imprecise or biased if the direction of causality runs opposite to that assumed in the \u0026nbsp;model - that is, if devolution and investment shifts are consequences, rather than causes, of regional economic growth. A large body of literature has sought to address the potential endogeneity of political decentralization and capital formation dynamics using instrumental variables. The model specified in Equation (1) includes two key variables\u0026mdash;devolution and investment changes\u0026mdash;that may be endogenous to growth, along with their interaction term, making identification strategies based on external instruments difficult to implement. Historical data on English regions, at the ITL3 level, \u0026nbsp;are also not easily attainable. To minimise endogeneity concerns, one approach would be to employ a dynamic panel analysis using a GMM-system estimator \u0026ndash; similar to that used by Calderon and Serven, 2004 and Crescenzi and Rodrıguez-Pose, 2012. Adopting a system-GMM estimator allows for a dynamic setup in controlling for unobserved heterogeneity and simultaneity while accommodating persistent growth effects. Put differently, this is a pragmatic approach used to circumvent the \u0026ldquo;instrument hunt\u0026rdquo; by relying on temporal dynamics.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe re-estimated model specified includes the two main variables of estimation: the devolution variable and change in investment proxy interacted with the devolution variable. Due to the number of instruments, the estimated model is limited to including only lags in GVA growth, changes in investment, the change in spatial investment lag, skills share, jobs density and population density. Industry shares are therefore excluded in the GMM estimation. In order to maintain a lower number of instruments than regions, second-order time lags are used as instruments.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe findings are in line with the baseline estimations. As per the GMM-system estimates, the devolution variable is consistently positive, but remains statistically insignificant. The interaction between devolution and changes in investment (proxied by total GFCF, intangible investment, buildings, \u0026lsquo;other tangibles\u0026rsquo;, ICT, and transport) show a positive relationship with GVA growth, but are also statistically insignificant.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis is not to say that such a methodology solves for endogeneity concerns, however. Devolution remains an exogenous policy measure, and the effectiveness of devolution is largely path-dependent, where time-lags may not necessarily allow for exogenous variation (Bun \u0026amp; Windmeijer, 2010). At best, employing such a check allows for validation of the base model estimations.\u0026nbsp;\u003c/p\u003e"},{"header":"4. Conclusions","content":"\u003cp\u003eUK governments, past and present, have prioritised devolution as a means to narrowing the productivity gap between the north and the south of the country, while allowing regions to have a bigger say in the prioritisation of the economic projects, initiatives, and objectives of respective city regions. Indeed, from the City Deals in the 2010s to the Growth Deals and \u0026apos;levelling-up\u0026apos; agenda that followed to the present version of devolution strategy (see Ministry of Housing, Communities, \u0026amp; Local Government, 2024), the role and emphasis on devolution in delivering better regional outcomes has never been greater. \u003cbr\u003e \u003cbr\u003e In the last 15 years, England has introduced over 10 combined authorities and metro mayors to streamline economic decision making so as to rejig the central-local framework that has existed for several decades. This paper attempts to shed some insights on the effectiveness of English devolution as currently constructed. The analysis presented should be seen as an early cross check of government devolution policy seeking to answer one fundamental question: has devolution and/or more devolved governance mechanisms led to better decision making, including and more specifically, better investment returns via better economic prioritisation of projects and investment initiatives.\u003cbr\u003e \u003cbr\u003e Crucially, the findings presented in the paper suggest no strong correlation between the level of devolution achieved by regions and returns to investment across England - be it in total gross fixed capital formation, ICT investment, intangible investment, \u0026apos;other tangibles\u0026apos;, and transport investment. Several robustness checks to cross check the baseline results were employed - all yielding similar results. Put simply, there is no positive and statistically significant link when it comes to devolution governance mechanisms in England and investment returns.\u003cbr\u003e \u003cbr\u003e The results uncovered adds to the empirical literature on English devolution in two ways. First, it is the only paper to our knowledge to codify regional governance mechanisms throughout the country. Second, it is also the first to make use of experimental regional investment statistics published by the ONS. In incorporating both elements, the paper adds to a scarce amount of quantitative research analysing the effects of devolution mechanisms and its impact on decision making and economic performance in the UK - something that has not been quantitatively investigated in much depth yet.\u003cbr\u003e \u003cbr\u003eTo be sure, this is not to say that devolution has failed, however, or yielded sub-optimal outcomes. The quantitative analysis presents many challenges and caveats, which should be carefully considered. For instance, investment (GFCF) data are still classified as \u0026apos;experimental\u0026apos; data. As with almost all national accounts statistics, data are prone to major revisions. Second, as the Institute for Government has highlighted, the effects of devolution may have yet to be fully felt, with some regions requiring years before we see any tangible differences in investment returns or indeed economic performance (Pope, et al., 2023). Specifically, given the time sample around which the analysis is based (2000-2019), any tangible policy effects may have yet to crystallise - particularly with infrastructure projects taking years for any benefits to be felt. Put simply, recent devolution policy (2010+) remains at its infancy and therefore may be harder to quantify at this juncture. A longer time-series of data may be needed to capture some of the devolution dividends associated with more effective economic prioritisation of projects and therefore growth. Moreover, changes in devolution are not random. Instead there is often political need for change \u0026ndash; and the last decade or so has been mired with economic malaise meaning that change in governance mechanism may initially coincide with weak economic performance. Lastly, the model used to assess the link between devolution and investment returns may suffer from omitted variable analysis, namely in the lack of control for institutional capacity and quality of governance \u0026ndash; which therefore raises further questions around central government strategy to push universally for devolution without tackling fundamental issues pertaining to the kind of devolution delivered across England.\u003cbr\u003e \u003cbr\u003eAs things stand, however, the analysis adds credence to the existing literature that devolution in the UK (and in England in particular) has failed to capture any tangible and genuine benefits given the still elevated role of Treasury in determining funding and investment decisions. Changes in governance mechanisms as seen in the last decade do not necessarily entail policy control. The Institute for Government notes that despite devolution being \u0026ldquo;a necessary precondition for stronger local institutions\u0026rdquo;, there are costs to devolution, including loss of economies of scale (resulting in an increase in administrative costs) and harmful competition (particularly with regards to the prevailing deal making funding framework employed centrally), see Pope, et al., 2023. Effective policy coordination depends on thorough analysis and strong stakeholder collaboration, necessitating skilled professionals. However, due to England\u0026rsquo;s centralized governance and the traditional clustering of civil servants in London, some regions face shortages of qualified policy experts (see Centre for Cities, 2024, Guerin, et al., 2021). \u003c/p\u003e\n\u003cp\u003eLocal/regional governments also lack sufficient spending power to deliver beyond their core functions, limiting the effectiveness of devolution mechanisms created in the last decade or two. As currently constructed, the two primary tax levers for English local government are council tax and business rates \u0026ndash; both of which comprise only around 25% of funding (well below the 40% average across other advanced economies (Haylen, 2019). Put simply, changes in the classification of sub-national government tiers do not imply effective change in policy making. The current funding for local/regional governments is essentially centred around competitively awarded grants. Local tax revenues play a very small role in funding day-to-day activities as well as investment. McCann (2022) highlights that the UK\u0026rsquo;s cost-based approach to funding local governments provides \u0026ldquo;much weaker fiscal stabilisation underpinnings than do revenue-based systems\u0026rdquo; with levels of subcentral government autonomy limiting local policy-making discretion while the UK\u0026rsquo;s over-centralised UK governance system \u0026ldquo;militates against both central government learning and local government institutional capacity building\u0026rdquo;. The Industrial Strategy Council too notes that subnational funding presented challenges given its competitive model with \u0026ldquo;stringent rules on spending along with ring-fenced budgets\u0026rdquo; limiting targeted, long-term interventions that would go a long way to address place-based issues while enabling longer-term strategic planning and implementation (Brittain \u0026amp; Taylor, 2021). McCann (2016) also concludes that \u0026ldquo;wholesale reform of the UK sub-central fiscal system must be a key part of an overall overhaul of the overly top-down and centralised UK governance system\u0026rdquo; with the \u0026ldquo;current UK central-subcentral fiscal system [representing] the worst of all worlds, in that it combines excessive centralisation with severe fragmentation\u0026rdquo;. In a recent report, the Resolution Foundation in partnership with the Centre for Cities (Breach \u0026amp; Bridgett, 2022) also found that \u0026ldquo;despite recent waves of devolution, Britain has become even more fiscally centralised since 2015, further reducing the already-weak incentive for local authorities to grow their local economies\u0026rdquo;. Providing better flexibility and medium-term funding certainty could go a long way in improving funding capabilities to deliver better investment returns. \u003c/p\u003e\n\u003cp\u003eLack of institutional capacity may have also hampered the abilities of combined (mayoral) authorities to reap any of the benefits of devolution. A dearth of local capacity could lead to uneven or less than optimal delivery. Pope et al. (2023) highlight the plight of mayoral authorities\u0026rsquo; staffing inadequacies, with places like Greater Manchester employing over 2,000 people while Cambridgeshire and Peterborough employ just over 50. The same report also highlighted that even in advanced mayoralties like greater Manchester and West Midlands, there are fewer than one employee per 1,000 inhabitants (compared to Toronto and Frankfurt who have more than 15 employees per 1,000 inhabitants). The IfG conclude that \u0026ldquo;England\u0026rsquo;s current local institutions are not suited to maximise the benefits of devolution\u0026rdquo;. Coyle and Muhtar (2021) also argue that the lack of cross government coordination often impedes the effectiveness of decentralisation/devolution, with asymmetric knowledge flows passing from the bottom to the top, as opposed to \u0026lsquo;top-down\u0026rsquo;. \u003c/p\u003e\n\u003cp\u003eThe way that devolution is currently structured also raises important questions in delivering better outcomes. Fragmentation horizontally and vertically have led to less effective governance mechanisms. Others have noted explicitly that fiscal devolution is made harder by the fact that local authorities are inconsistent with local economies (see Breach \u0026amp; Bridgett, 2022, McCann, 2022). Local services are managed at various levels of political decentralisation, further complicating the UK\u0026rsquo;s devolution progress. When considering the current state of combined authorities, the Productivity Institute (Shaw, 2024) notes that the \u0026ldquo;smaller scale of combined authorities is also likely to create institutions with insufficient capacity and capability, as well as \u0026lsquo;weaker voices\u0026rsquo; that are more easily ignored by government.\u0026rdquo; The lack of effective sub-national governance means that locally specific \u0026apos;knowledge inputs\u0026apos; fail to reach government decision-makers (Coyle \u0026amp; Muhtar, 2021). Put simply, governance mechanisms \u0026ndash; as currently structured \u0026ndash; do little to improve regional leadership whilst also reducing local and regional accountability. \u003c/p\u003e\n\u003cp\u003eSeveral conclusions can be taken from the above analysis. While devolution in theory entails many potential benefits, it is also apparent that its effects have not been dramatic in improving investment decision making across England. A shift to flexible funding arrangements may better able combined (mayoral) authorities to deliver on local objectives and growth. A shift away from away from competitive funding pots may also prove more effective in delivering longer-lasting economic transformation \u0026ndash; one that takes time and one that requires long-term regional strategic vision. The ability to hire and train staff to build institutional capacity to improve governance quality may also be a necessary pre-condition for better economic returns from devolution. Increased devolution structures may also need to be better structured within the central-local establishment, whereby Treasury enables local government to have more autonomy in setting regional economic strategies, whilst also working with other authorities to create a more holistic meso-regional development policy agenda \u0026ndash; one that effectively works to narrow the North-South divide and improve regional productivity rather than merely creating independent economic spatial silos. This paper should serve as a reminder for policy makers that devolution in and of itself is not a panacea to deliver better economic performance and improve regional outcomes. The lack of independent evaluation of present spatial policies, including and perhaps most importantly, devolution, raises concerns around past and present governments\u0026rsquo; rush to devolve \u0026ndash; with many seeing devolution as a \u0026lsquo;quick fix\u0026rsquo; to treat all regional economic ailments. Getting existing devolution right should be the higher priority rather than rushing towards a broad-scale devolution policy that proves piecemeal in nature and ultimate is less effective \u0026ndash; potentially even exacerbating spatial inequality in England.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eSR wrote the main manuscript text, figures, and implemented all the analyses.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThis paper has greatly benefitted from comments and suggestions by Dr Johan Larsson and Professor Diane Coyle. All errors are my own.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eI have used publicly available data to generate a time series of the English political devolution variable.\u003c/p\u003e\u003ch2\u003e\u003cstrong\u003e\u003cem\u003eFunding Declaration:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003e\u003cem\u003eNo funding was received for this research.\u0026nbsp;\u003c/em\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBurret , H. T., Feld, L. P. \u0026amp; Schaltegger, C. A., 2022. Fiscal federalism and economic performance new evidence from Switzerland. \u003cem\u003eEuropean Journal of Political Economy, \u003c/em\u003e74(102159).\u003c/li\u003e\n\u003cli\u003eAcemoglu, , D. \u0026amp; Dell, M., 2010. Productivity Differences between and within Countries. \u003cem\u003eAmerican Economic Journal: Macroeconomics, \u003c/em\u003eVolume 2, pp. 169-188.\u003c/li\u003e\n\u003cli\u003eAcemoglu, D., 2002. Technical Change, Inequality, and the Labor Market. \u003cem\u003eJournal of Economic Literature, \u003c/em\u003e40(1), pp. 7-72.\u003c/li\u003e\n\u003cli\u003eAschauer, D. A., 1989. Is public expenditure productive?. \u003cem\u003eJournal of Monetary Economics, \u003c/em\u003e23(2), pp. 177-200.\u003c/li\u003e\n\u003cli\u003eAutor, , D., Levy, F. \u0026amp; Murnane, R. J., 2003. The Skill Content of Recent Technological Change: An Empirical Exploration. \u003cem\u003eQuarterly Journal of Economics, \u003c/em\u003e118(4), pp. 1279-1333.\u003c/li\u003e\n\u003cli\u003eBanister, D. \u0026amp; Berechman, J., 2003. \u003cem\u003eTransport Investment and Economic Development. \u003c/em\u003e1st ed. London: Routledge.\u003c/li\u003e\n\u003cli\u003eBardhan, P. \u0026amp; Mookherjee, D., 2006. Decentralization, Corruption and Government Accountability. In: S. Rose-Ackerman, ed. \u003cem\u003eInternational Handbook on the Economics of Corruption. \u003c/em\u003es.l.:Edward Elgar Publishing.\u003c/li\u003e\n\u003cli\u003eBenhabib, . J. \u0026amp; Spiegel, M. M., 1994. The Role of Human Capital in Economic Development: Evidence from Aggregate Cross-Country Data.. \u003cem\u003eJournal of Monetary Economics, \u003c/em\u003e34(2), pp. 143-173.\u003c/li\u003e\n\u003cli\u003eBeveridge, S. W., 1942. \u003cem\u003eSocial Insurance and Allied Services. \u003c/em\u003eLondon: His Majesty\u0026apos;s Stationary Office.\u003c/li\u003e\n\u003cli\u003eBl\u0026ouml;chliger, H. \u0026amp; \u0026Eacute;gert, . B., 2013. \u003cem\u003eDecentralisation and Economic Growth - Part 2: The Impact on Economic Activity, Productivity and Investment, \u003c/em\u003eParis: OECD Publishing.\u003c/li\u003e\n\u003cli\u003eBloom, N. et al., 2013. Does Management Matter? Evidence from India. \u003cem\u003eThe Quarterly Journal of Economics, \u003c/em\u003e128(1), pp. 1-51.\u003c/li\u003e\n\u003cli\u003eBreach, A. \u0026amp; Bridgett, S., 2022. \u003cem\u003eCentralisation Nation, \u003c/em\u003eLondon: Resolution Foundation.\u003c/li\u003e\n\u003cli\u003eBrittain, B. \u0026amp; Taylor, A., 2021. \u003cem\u003eDevolution and Governance Structures in the UK, \u003c/em\u003es.l.: West Midlands Regional Economic Development Institute.\u003c/li\u003e\n\u003cli\u003eBritteon, P. et al., 2024. The impact of devolution on local health systems: Evidence from Greater Manchester, England. \u003cem\u003eSocial Science \u0026amp; Medicine, \u003c/em\u003e348(116801).\u003c/li\u003e\n\u003cli\u003eBun, M. J. G. \u0026amp; Windmeijer, F., 2010. The weak instrument problem of the system GMM estimator in dynamic panel data models. \u003cem\u003eThe Econometrics Journal, \u003c/em\u003e13(1), pp. 95-126.\u003c/li\u003e\n\u003cli\u003eCalderon, C. \u0026amp; Serven, L., 2004. \u003cem\u003eThe Effects of Infrastructure Development on Growth and Income, \u003c/em\u003eWashington, D.C.: World Bank Policy Research Working Paper.\u003c/li\u003e\n\u003cli\u003eCentre for Cities, 2024. \u003cem\u003eCities Outlook 2024, \u003c/em\u003eLondon: Centre for Cities.\u003c/li\u003e\n\u003cli\u003eCiccone, A. \u0026amp; Hall, R. E., 1996. Productivity and the Density of Economic Activity. \u003cem\u003eAmerican Economic Review, \u003c/em\u003e86(1), pp. 54-70.\u003c/li\u003e\n\u003cli\u003eCoyle , D. \u0026amp; Muhtar, A., 2021. \u003cem\u003eUK\u0026rsquo;s Industrial Policy: Learning from the past?, \u003c/em\u003eCambridge: University of Cambridge - Bennett Institute for Public Policy.\u003c/li\u003e\n\u003cli\u003eCoyle, D. \u0026amp; Hampton, L., 2023. \u003cem\u003eTwenty-first century progress in computing, \u003c/em\u003eCambridge: Bennett Institute for Public Policy, University of Cambridge.\u003c/li\u003e\n\u003cli\u003eCrain, M. \u0026amp; Oakley, L. K., 1995. The Politics of Infrastructure. \u003cem\u003eJournal of Law and Economics, \u003c/em\u003eVolume 38, pp. 1-17.\u003c/li\u003e\n\u003cli\u003eCrescenzi, R., Di Cataldo, M. \u0026amp; Rodrıguez-Pose, A., 2016. Government Quality and the Economic Returns of Transport Infrastructure Investment in European Regions. \u003cem\u003eJournal of Regional Science, \u003c/em\u003e56(4), pp. 555-582.\u003c/li\u003e\n\u003cli\u003eCrescenzi, R. \u0026amp; Rodrıguez-Pose, A., 2012. Infrastructure and Regional Growth in the European Union. \u003cem\u003ePapers in Regional Science, \u003c/em\u003eVolume 91, pp. 487-513.\u003c/li\u003e\n\u003cli\u003eEasterly, W. R., 2002. \u003cem\u003eThe Elusive Quest for Growth: Economists\u0026apos; Adventures and Misadventures in the Tropics. \u003c/em\u003e2nd ed. Cambridge: MIT Press.\u003c/li\u003e\n\u003cli\u003eEnikolopov, R. \u0026amp; Zhuravskaya, E., 2007. Decentralization and political institutions. \u003cem\u003eJournal of Public Economics, \u003c/em\u003e91(11-12), pp. 2261-2290.\u003c/li\u003e\n\u003cli\u003eFaguet , J.-P. \u0026amp; P\u0026ouml;schl , C., 2015. Is decentralization good for development? Perspectives from academics and policy makers. In: Oxford: Oxford University Press, pp. 1-29.\u003c/li\u003e\n\u003cli\u003eFaguet, J.-P., 2021. \u003cem\u003eUnderstanding decentralization: theory, evidence and method, with a focus on least-developed countries., \u003c/em\u003eLondon: LSE Working Paper.\u003c/li\u003e\n\u003cli\u003eFisman, R. \u0026amp; Gatti, R., 2002. Decentralization and corruption: evidence across countries. \u003cem\u003eJournal of Public Economics, \u003c/em\u003e83(3), pp. 325-345.\u003c/li\u003e\n\u003cli\u003eGennaioli, N., La Porta, R. \u0026amp; Lopez-de-Silanes, F., 2013. Human Capital and Regional Development.. \u003cem\u003eQuarterly Journal of Economics, \u003c/em\u003e128(1), pp. 105-164.\u003c/li\u003e\n\u003cli\u003eGLA Economics, 2022. \u003cem\u003eDevolution and economic growth in London: A rapid evidence review. , \u003c/em\u003eLondon: Greater London Authority.\u003c/li\u003e\n\u003cli\u003eGlaeser, E. \u0026amp; Resseger, M. G., 2010. The Complementarity Between Cities and Skills.. \u003cem\u003eJournal of Regional Science, \u003c/em\u003e50(1), pp. 221-244.\u003c/li\u003e\n\u003cli\u003eGuerin, B., Thomas, A., Clyne, R. \u0026amp; Vira, S., 2021. \u003cem\u003eFinding the right skills for the civil service, \u003c/em\u003eLondon: Institute for Government.\u003c/li\u003e\n\u003cli\u003eHanushek, E. A., , E. A. \u0026amp; Kimko, D., 2000. Schooling, Labor-Force Quality, and the Growth of Nations.. \u003cem\u003eAmerican Economic Review, \u003c/em\u003e90(5), pp. 1184-1204.\u003c/li\u003e\n\u003cli\u003eHaylen, A., 2019. \u003cem\u003eLocal road maintenance in England, \u003c/em\u003eLondon: House of Commons Library.\u003c/li\u003e\n\u003cli\u003eHenisz,, W. J., 2002. The Institutional Environment for Infrastructure Investment. \u003cem\u003eIndustrial and Corporate Change, \u003c/em\u003eVolume 11, pp. 355-389.\u003c/li\u003e\n\u003cli\u003eHirshman, O., 1970. \u003cem\u003eExit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. \u003c/em\u003eCambridge: Harvard University Press.\u003c/li\u003e\n\u003cli\u003eHitchcock, A., Borrows, M. \u0026amp; Harwich, E., 2017. \u003cem\u003eVive la devolution: devolved public-services commissioning, \u003c/em\u003eLondon: Reform.\u003c/li\u003e\n\u003cli\u003eLaverty, K. J., 1996. Economic \u0026quot;Short-Termism\u0026quot;: The Debate, the Unresolved Issues, and the Implications for Management Practice and Research. \u003cem\u003eThe Academy of Management Review, \u003c/em\u003e21(3), pp. 825-860.\u003c/li\u003e\n\u003cli\u003eMcCann, P., 2022. \u003cem\u003eThe Fiscal Implications of \u0026apos;Levelling Up\u0026apos; and UK Governance Devolution, \u003c/em\u003eLondon: NIESR Occasional Paper LXIII.\u003c/li\u003e\n\u003cli\u003eMeloche, J.-P., Vaillancourt, F. \u0026amp; Yilmaz, S., 2004. \u003cem\u003eDecentralization or Fiscal Autonomy? What Does Really Matter? Effects on Growth and Public Sector Size in European Transition Countries., \u003c/em\u003eWashington D.C.: World Bank.\u003c/li\u003e\n\u003cli\u003eMinistry of Housing, Communities, \u0026amp; Local Government, 2024. \u003cem\u003eEnglish Devolution White Paper, \u003c/em\u003eLondon: UK Government.\u003c/li\u003e\n\u003cli\u003eMuringani, J., Fitjar, R. D. \u0026amp; Rodr\u0026iacute;guez-Pose, A., 2019. \u003cem\u003eDecentralisation, quality of government and economic growth in the regions of the EU, \u003c/em\u003eLondon: LSE Research Online.\u003c/li\u003e\n\u003cli\u003eOates, W. E., 1973. Fiscal Federalism. \u003cem\u003ePublic Choice, \u003c/em\u003eVolume 14, pp. 155-157.\u003c/li\u003e\n\u003cli\u003eOECD, 2019. \u003cem\u003eMaking Decentralisation Work: A Handbook for Policy-Makers, \u003c/em\u003eParis: OECD Publishing.\u003c/li\u003e\n\u003cli\u003ePope, T., Dalton, G. \u0026amp; Coggins, M., 2023. \u003cem\u003eHow can devolution deliver regional growth in England?, \u003c/em\u003eLondon: Institute for Government.\u003c/li\u003e\n\u003cli\u003eRaja, S. I. \u0026amp; Larsson, J. P., 2024. Have City Deals delivered higher productivity in England? An empirical assessment of a broad-spectrum local growth policy.. \u003cem\u003eUrban Studies.\u003c/em\u003e\u003c/li\u003e\n\u003cli\u003eRedding, S. J. \u0026amp; Turner, M. A., 2015. Transportation Costs and the Spatial Organization of Economic Activity. \u003cem\u003eHandbook of Regional and Urban Economics , \u003c/em\u003eVolume 5, pp. 1339-1398.\u003c/li\u003e\n\u003cli\u003eRodr\u0026iacute;guez-Pose, A. \u0026amp; Ezcurra, R., 2010. Does decentralization matter for regional disparities? A cross-country analysis. \u003cem\u003eJournal of Economic Geography, \u003c/em\u003e10(5), pp. 619-644.\u003c/li\u003e\n\u003cli\u003eRodr\u0026iacute;guez-Pose, A. \u0026amp; Ezcurra, R., 2011. Is fiscal decentralization harmful for economic growth? Evidence from the OECD countries. \u003cem\u003eJournal of Economic Geography, \u003c/em\u003e11(4), pp. 619-643.\u003c/li\u003e\n\u003cli\u003eRodr\u0026iacute;guez‐Pose, A. \u0026amp; Gill, N., 2005. On the \u0026lsquo;economic dividend\u0026rsquo; of devolution. \u003cem\u003eRegional Studies, \u003c/em\u003e39(4), pp. 405-420.\u003c/li\u003e\n\u003cli\u003eShaw, J., 2024. \u003cem\u003eDevolution: the importance of scale and coterminosity, \u003c/em\u003eLondon: The Productivity Institute.\u003c/li\u003e\n\u003cli\u003eSow, M. \u0026amp; Razafimahefa, I. F., 2015. \u003cem\u003eFiscal Decentralization and the Efficiency of Public Service Delivery, \u003c/em\u003eWashington D.C.: International Monetary Fund.\u003c/li\u003e\n\u003cli\u003eTiebout, C. M., 1956. A Pure Theory of Local Expenditures. \u003cem\u003eJournal of Political Economy, \u003c/em\u003e64(5), pp. 416-424.\u003c/li\u003e\n\u003cli\u003eTomaney, J., 2016. Limits of Devolution: Localism, Economics and Post-democracy. \u003cem\u003eThe Political Quarterly, \u003c/em\u003eVolume 87.\u003c/li\u003e\n\u003cli\u003eTorrance, D., 2019. \u003cem\u003eIntroduction to devolution in the UK, \u003c/em\u003eLondon: House of Commons Library.\u003c/li\u003e\n\u003cli\u003eTreisman, D., 2007. \u003cem\u003eThe Architecture of Government: Rethinking Political Decentralization. \u003c/em\u003eCambridge: Cambridge University Press.\u003c/li\u003e\n\u003cli\u003eUK Government, 2015. \u003cem\u003e2010 to 2015 government policy: UK economic growth, \u003c/em\u003eLondon: House of Commons.\u003c/li\u003e\n\u003cli\u003eZhao, C., Feng, F., Chen, Y. \u0026amp; Li, X., 2021. Local government competition and regional innovation efficiency: From the perspective of China-style fiscal federalism. \u003cem\u003eScience and Public Policy, \u003c/em\u003e48(4), pp. 488-489. \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"the-annals-of-regional-science","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"arsc","sideBox":"Learn more about [The Annals of Regional Science](https://link.springer.com/journal/168)","snPcode":"168","submissionUrl":"https://submission.springernature.com/new-submission/168/3","title":"The Annals of Regional Science","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Devolution, investment, regional governance, economic development, decentralisation, spatial policy, United Kingdom","lastPublishedDoi":"10.21203/rs.3.rs-8896200/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8896200/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Despite the rapid expansion of devolution in England over the past two decades, with local andregional governments gaining unprecedented control over economic policy levers, the anticipatedeconomic dividends remain elusive. Using novel data on regional gross capital formation from 2000to 2019, this paper delivers one of the first quantitative assessments of the relationship between thescale of devolution and investment returns across key sectors. Contrary to prevailing policy narrativesand international evidence suggesting decentralisation boosts economic performance, our findingsreveal no significant link between increased devolution and improved investment returns in England.Robustness checks reinforce this outcome, highlighting persistent challenges such as limited fiscalautonomy, institutional capacity constraints, and ill-defined economic geographies. These results callinto question the effectiveness of England’s current devolution model in driving better economicprioritisation and underscore the urgent need for further research into the mechanisms required tounlock the full potential of place-based policy reforms.","manuscriptTitle":"Has Political Devolution Led to Better Investment Returns? A Look at English Governance Mechanisms Over The Last Two Decades","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-05 18:33:32","doi":"10.21203/rs.3.rs-8896200/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewersInvited","content":"","date":"2026-04-24T07:59:38+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-02-23T11:03:47+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-02-23T10:58:45+00:00","index":"","fulltext":""},{"type":"submitted","content":"The Annals of Regional Science","date":"2026-02-16T20:51:31+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"the-annals-of-regional-science","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"arsc","sideBox":"Learn more about [The Annals of Regional Science](https://link.springer.com/journal/168)","snPcode":"168","submissionUrl":"https://submission.springernature.com/new-submission/168/3","title":"The Annals of Regional Science","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"6b108e4f-f488-427a-95ab-0bf5251bf898","owner":[],"postedDate":"May 5th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-05T18:33:32+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-05 18:33:32","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8896200","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8896200","identity":"rs-8896200","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}