The Paradox of Health Financing in Morocco: An Exploratory Analysis using a VAR Model (2000-2022)

The Paradox of Health Financing in Morocco: An Exploratory Analysis using a VAR Model (2000-2022) | 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 The Paradox of Health Financing in Morocco: An Exploratory Analysis using a VAR Model (2000-2022) Aazelarab BOUGHALEB, JERRY Mounir This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8106278/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Despite a spectacular expansion of its insurance coverage, Morocco faces a persistent paradox: the share of out-of-pocket (OOP) expenditures in health financing remains high, standing at around 43% in 2022. This article aims to explore the macroeconomic dynamics of this paradox by analyzing the interactions between public health spending, OOP, and total health expenditure. To do this, a Vector Autoregressive (VAR) model is applied to time series from the World Bank and WHO covering the period 2000–2022. The results do not reveal statistically significant Granger causality between changes in public spending and those in OOP. The impulse response function analysis, while showing a negative response of OOP to a public spending shock, confirms the lack of statistical significance. The variance decomposition highlights a strong inertia, with over 71% of OOP fluctuations explained by their own past dynamics. These findings suggest that the relationship between public financing efforts and the alleviation of the household burden is neither direct nor automatic at the aggregate level. The study concludes on the limitations of macroeconomic analysis in capturing this complex phenomenon and underscores the imperative need to resort to micro-econometric analyses to identify the real determinants of financial risk at the household level. Health financing Out-of-Pocket expenditures (OOP) Universal Health Coverage (UHC) Financial protection VAR model Time series Figures Figure 1 Figure 2 Figure 3 Figure 4 INTRODUCTION The path towards Universal Health Coverage (UHC) is a major objective of global health policies, enshrined as a central target of the Sustainable Development Goals [ 1 ]. UHC is based on a dual imperative: ensuring equitable access to quality health services for the entire population and providing effective financial protection against the costs of illness [ 2 ]. Among the obstacles to this protection, out-of-pocket (OOP) expenditures are identified as the most inequitable mode of financing and the main driver of catastrophic health spending, capable of plunging families into poverty [ 3 ]. Reducing the reliance of health systems on OOP is therefore a key marker of progress towards UHC. In this global context, Morocco presents a particularly paradoxical case study. Since the beginning of the 21st century, the Kingdom has undertaken a series of ambitious reforms to expand health coverage, culminating in the royal project for the generalization of social protection launched in 2021 [ 4 ]. This effort has led to a spectacular expansion of formal insurance coverage, from less than 20% of the population in the early 2000s to a goal of near-universal coverage [ 5 ]. However, this progress confronts a persistent reality: despite a substantial increase in the share of public and insurance financing, the share of out-of-pocket expenditures in total health financing remains structurally high. In 2022, OOP still accounted for 43% of current health expenditures, a level well above the 15–20% threshold considered a "danger zone" by the World Health Organization [ 6 ]. This disconnect between the extension of de jure coverage and the improvement of de facto financial protection constitutes the "Moroccan paradox." While the literature has well-documented the extent of this burden [ 7 ], few quantitative studies have sought to analyze the macroeconomic dynamics of this paradox over the long term. How have the interactions between public financing efforts and household spending evolved throughout the major reforms? Have public policies had a measurable and significant impact on alleviating the citizens' burden? This article aims to provide a first econometric insight into these questions. The objective is to explore the dynamic interactions between public health spending, out-of-pocket expenditures, and total health expenditure in Morocco over the period 2000–2022. By mobilizing a Vector Autoregressive (VAR) model, this study seeks to characterize the nature of the dynamic relationship between these main components of financing. The goal is less to prove a definitive causality than to analyze the propagation of shocks, test for predictability, and evaluate the relative strength of mutual influences, in order to better understand the system's inertias and levers. This article is organized as follows. The next section presents the methodology and the data used. The third section presents the results of the VAR analysis, including stationarity tests, model selection, Granger causality tests, impulse response functions, and variance decomposition. Finally, the fourth section discusses the implications of these macroeconomic results and highlights the limitations of this approach, advocating for complementary micro-econometric analyses for a comprehensive understanding of the phenomenon. METHODOLOGY AND DATA To analyze the dynamic interactions between the different components of health financing in Morocco, this exploratory study adopts an econometric approach based on time series. The strategy consists of modeling the relationships between key variables using a Vector Autoregressive (VAR) model, a method particularly suited for capturing dynamic interdependencies without imposing strong a priori theoretical constraints [ 8 ]. Data: The data used in this study come from the World Health Organization's Global Health Expenditure Database, accessed via the World Bank's Open Data platform [ 6 ]. The analysis period spans 23 years, from 2000 to 2022, providing a sufficient chronological series for a dynamic analysis. Three key variables were selected to build our model: Out-of-pocket expenditures (oop_pct) : Measured as a percentage of current health expenditures. This variable represents the direct financial burden borne by the population. Government health expenditures (gouv_pct) : Measured as a percentage of current health expenditures. This variable represents the collective financing effort of the state and social security schemes. Total health expenditures (dts_pib_pct) : Measured as a percentage of Gross Domestic Product (GDP). This variable serves as a control variable to capture the overall evolution of the health sector's importance in the national economy. Econometric Strategy: The estimation of a VAR model requires that the time series be stationary. A series is said to be stationary if its mean, variance, and autocovariance are constant over time. Most macroeconomic series exhibit trends and are therefore non-stationary in levels, which can lead to spurious regressions [ 9 ]. The econometric approach therefore followed these steps: Stationarity analysis : The stationarity of each series was tested using the Augmented Dickey-Fuller (ADF) test [ 10 ]. The null hypothesis (H₀) of this test is the presence of a unit root, i.e., non-stationarity of the series. If the p-value of the test is greater than the 5% significance level, the null hypothesis cannot be rejected. In this case, the series is differenced (by calculating its change from one period to the next) and the test is reapplied until stationarity is achieved. Selection of the optimal number of lags (p) : A VAR(p) model expresses each variable as a linear function of its own past values and the past values of all other variables in the system, up to a lag p. The choice of p is crucial: a p that is too small may omit important dynamics, while a p that is too large consumes degrees of freedom and may overfit the model. The optimal number of lags was determined using several sequential information criteria, notably the Akaike Information Criterion (AIC) and the Schwarz Bayesian Information Criterion (SC) [ 11 ]. VAR model estimation : Once the number of lags p was selected and the stationarity of the variables was ensured, the VAR model was estimated using the Ordinary Least Squares (OLS) method. Analysis of results : Three main tools derived from the estimated VAR model were used for the analysis: Granger causality tests : To determine if the past values of one variable help predict the future values of another variable [ 12 ]. Impulse Response Functions (IRF) : To trace the path of the effect of a "shock" (an unexpected innovation) in one variable on the future evolution of the other variables in the system [ 13 ]. Forecast Error Variance Decomposition (FEVD) : To quantify the proportion of the forecast error variance of each variable attributable to shocks in the other variables [ 13 ]. All analyses were performed using R software (version 4.5.1) and the vars package [ 14 ]. RESULTS The application of the methodology described above allowed for the characterization of the health financing dynamics in Morocco. This section successively presents the results of the preliminary tests, model selection, and then the analyses of causality, impulse response, and variance decomposition. Stationarity and model selection The plot of the time series in levels (Fig. 1) suggests the presence of trends, particularly for public expenditures (gouv_pct) and total expenditures as a percentage of GDP (dts_pib_pct). Formal stationarity tests confirm this visual observation. Table 1 presents the results of the Augmented Dickey-Fuller (ADF) tests. The null hypothesis of non-stationarity (presence of a unit root) is rejected at the 5% significance level for the oop_pct series (p = 0.045), indicating that it is stationary in levels, I(0). In contrast, this hypothesis is not rejected for the gouv_pct (p = 0.247) and dts_pib_pct (p = 0.636) series, which are therefore considered non-stationary, I(1). To build a standard VAR model, all variables must have the same order of integration. Consequently, all series were differenced once. The ADF tests re-applied to the first-differenced series (Δ) confirmed their stationarity (results available upon request). Figure 2 shows the evolution of these differenced series, which now oscillate around zero. The choice of the optimal number of lags (p) for the VAR model was guided by the information criteria presented in Table 2. The Akaike criterion (AIC), which is often favored for small samples, reaches its minimum for a lag of p = 1. Therefore, a VAR(1) model was estimated on the first-differenced data. Diagnostic tests of the model (autocorrelation, normality, heteroscedasticity) confirmed its good specification. Granger causality analysis Table 3 presents the results of the Granger causality tests. The null hypothesis is the absence of causality from the "cause" variable to the other variables in the system. For the two relationships of interest, the p-value is well above the 5% threshold. Thus, over the period studied, past variations in public spending do not provide statistically significant information to predict future variations in out-of-pocket expenditures (p = 0.306). Likewise, the inverse causality is also not established (p = 0.349). This absence of Granger causality suggests that the links between the annual variations of these two aggregates are complex and do not follow a simple predictive relationship, a conclusion that can be partly attributed to the low statistical power due to the limited sample size. Impulse Response Analysis Figure 3 shows the impulse response functions (IRF) of the variation in OOP (oop_pct) and total health expenditures (dts_pib_pct) following an exogenous one-standard-deviation shock to the variation in public expenditures (gouv_pct). The average response of the OOP variation (upper panel, black line) is negative. Immediately after the shock, the OOP variation decreases, reaching its lowest point (about − 0.2 percentage points) one year after the shock, before gradually returning to zero. This trajectory is consistent with economic theory: an unexpected increase in public effort tends to reduce (or slow the growth of) the household burden. However, the 95% confidence interval (red dashed lines) includes the zero line over the entire forecast horizon. This means that, although the direction of the effect is as expected, the impact is not statistically significant at the 5% level. Forecast Error Variance Decomposition (FEVD) The variance decomposition analysis allows for the quantification of the relative contribution of each variable to the unforeseen fluctuations of the others. Table 4 details the share of the forecast error variance of the oop_pct variable explained by shocks from each variable in the system, at different time horizons. The results are unequivocal. In the short term (1-year horizon), 100% of the unexpected fluctuations in OOP are explained by its own past shocks. Even at a longer horizon of 5 years, this share remains predominant at 71.72%. Shocks to public spending (Shocks on Gov. Exp.) explain only a very marginal part of the OOP variance, reaching just 2.22% at the 5-year horizon. In contrast, shocks to total health expenditure as a percentage of GDP (Shocks on THE (%GDP)) explain a more substantial share, reaching 26.06%. These results indicate a strong inertia in the dynamics of out-of-pocket expenditures, whose variations are primarily determined by their own history, and a low sensitivity to annual budgetary policy shocks over the period studied. DISCUSSION The results of the econometric analysis using a VAR model, though conducted over a period of more than two decades, present a nuanced and complex picture of health financing dynamics in Morocco. Three main findings emerge from our analyses: a strong inertia of out-of-pocket expenditures, an absence of statistically significant Granger causality, and a negative but non-significant response of OOP to public financing shocks. Far from invalidating the relevance of the issue, these results call for a deeper interpretation of the mechanisms at play and the limitations of macroeconomic analysis. The most salient result is the strong inertia of the share of out-of-pocket expenditures. The variance decomposition (Table 4) showed that more than 71% of unexpected fluctuations in OOP at a five-year horizon are explained by its own past shocks. This structural inertia suggests that the level of OOP is less sensitive to cyclical annual variations than to deep-seated factors, profoundly anchored in the architecture of the health system. This echoes the analysis of the historical trajectory of reforms, which highlighted the progressive constitution of a dual system where recourse to the private sector and direct purchase of medical goods have become firmly established practices for a large segment of the population, including the insured [ 5 ]. This "path dependence" creates a dynamic where the behaviors of households and providers are difficult to influence through simple annual budgetary adjustments. Secondly, the absence of statistically significant Granger causality between variations in public spending and those in OOP (Table 3) must be interpreted with caution. Econometrically, this means that, based on the available data, the past variations of one series do not significantly improve the prediction of the future variations of the other. Economically, this does not mean an absence of a relationship, but rather that the link is neither simple, nor direct, nor immediate. This conclusion is reinforced by the analysis of the impulse response functions (Fig. 3). Although the average response of OOP to a positive shock on public spending is negative, as expected, the effect is not statistically distinct from zero. Several contextual factors can explain this apparent statistical disconnect: Statistical limitations : The sample size (23 annual observations) is small for a time series analysis, which considerably reduces the power of statistical tests. It is likely that a real economic relationship exists, but that the sample is not large enough to detect it with a 95% confidence level [ 9 ]. Allocation of public spending : The increase in public health spending was not necessarily allocated to measures directly aimed at reducing out-of-pocket costs. A significant portion may have been absorbed by salary increases, the construction of new infrastructure, or the acquisition of expensive technologies, without proportionally translating into a lighter burden for households [ 7 ]. Cost-inflation effect in the private sector : It is possible that the expansion of insurance coverage created a windfall effect in a poorly regulated private sector, leading to an increase in tariffs and balance billing that would have "absorbed" part of the expected gains from the public effort [ 15 ]. These macroeconomic results strongly underscore the need to move to a micro-level analysis. The VAR model, by construction, deals with national averages that mask very diverse realities. It cannot capture the glaring inequalities between households, nor the coping strategies, nor the impact of socio-demographic characteristics (age, chronic illness, place of residence) which are nevertheless the central determinants of the risk of catastrophic expenditure [ 3 , 16 ]. The non-significance of the aggregate results therefore only reinforces the relevance of the initial research question: if the macro dynamics do not provide a clear answer, it is at the level of households and system actors that we must seek the keys to "open the black box" of the Moroccan paradox. CONCLUSION This study aimed to explore the macroeconomic dynamics of the "Moroccan paradox" by analyzing the interactions between public health spending and out-of-pocket expenditures over the period 2000–2022. By mobilizing a Vector Autoregressive (VAR) model, we sought to characterize the nature of this complex relationship throughout the successive reforms of the health system. The results reveal a nuanced picture. We found no statistically significant evidence of Granger causality between annual variations in public financing and those in out-of-pocket expenditures. Similarly, while the impulse response analysis suggests that a positive shock on public spending tends to reduce household expenditures, this effect is not statistically distinct from zero. The main takeaway from our model is the strong inertia in the dynamics of out-of-pocket expenditures, whose fluctuations are overwhelmingly explained by their own history, and a low sensitivity to annual budgetary policy shocks [Table 4]. The main implication of these results is both methodological and substantive. From a methodological standpoint, they highlight the limits of purely macroeconomic analysis for grasping the concrete effects of health policies, particularly with short time series. From a substantive standpoint, the absence of a strong statistical relationship suggests that the link between public investment and the financial protection of households is neither direct nor automatic. This inertia points to the preponderant influence of structural factors—such as the organization of the care market, weaknesses in tariff regulation, and actor behaviors—which seem to hamper the full translation of public effort into an alleviation of the household burden [ 15 ]. This study has limitations, the main one being the short length of the time series, which affects the power of the econometric tests. Furthermore, the use of aggregated data inherently masks the strong heterogeneities existing among households. These limitations, however, trace a clear path for future research. To truly "open the black box" of the Moroccan paradox, it is now imperative to change the scale of analysis. Research must move from the macroeconomic to the microeconomic, relying on household survey data. It is at this level that the real determinants of catastrophic spending risk can be robustly identified [ 16 ]. This article, by providing a first quantitative characterization of the health financing dynamics in Morocco, demonstrates that while public effort in favor of health is a necessary condition, it is not sufficient on its own to guarantee the financial protection of citizens. The resolution of the paradox of high out-of-pocket expenditures lies not only in the volume of financing, but in deep structural reforms of the health system and a fine-grained understanding of microeconomic factors, which constitutes a priority research area. Declarations Ethical Approval Ethical approval was not required for this study as it is based exclusively on the analysis of publicly available, aggregated, and anonymized time-series data from the World Bank and the World Health Organization (Global Health Expenditure Database). The research did not involve human participants or access to individual-level confidential data. Conflict of Interests The authors declare that they have no competing interests to disclose. Funding The authors declare that they received no specific funding for this research. Author Contribution A.B.: Conceptualization, Methodology, Software, Data Curation, Formal Analysis, Investigation, Writing – Original Draft. M.J.: Supervision, Validation, Writing – Review & Editing. All authors have read and agreed to the published version of the manuscript. Data Availability The datasets analyzed during the current study are publicly available in the World Bank Open Data repository. The data originates from the World Health Organization (WHO) Global Health Expenditure database and can be accessed via the following link: https://data.worldbank.org/indicator/SH.XPD.OOPC.CH.ZS?locations=MA. References United Nations. Transforming our world: the 2030 Agenda for Sustainable Development. General Assembly (2015) Resolution A/RES/70/1 World Health Organization (2010) Health systems financing: the path to universal coverage. World Health Report 2010. WHO, Geneva Xu K, Evans DB, Carrin G, Aguilar-Rivera AM, Musgrove P, Evans T (2007) Protecting households from catastrophic health spending. Health Aff 26(4):972–983 Kingdom of Morocco (2021) Framework Law 09–21 on social protection. Official Gaz No 6975; April 5 Economic, Social and Environmental Council (2024) Généralisation de l'AMO, bilan d'étape: Une avancée sociale à consolider, des défis à relever [Generalization of Compulsory Health Insurance, progress report: A social advance to be consolidated, challenges to be met]. Opinion No. 80/2024. CESE, Rabat World Bank Global Health Expenditure database [Internet]. World Bank Open Data. [cited June 1, 2025]. Available from: https://data.worldbank.org/indicator/SH.XPD.OOPC.CH.ZS?locations=MA Ministry of Health and Social Protection (2022) Comptes Nationaux de la Santé 2022 [National Health Accounts 2022]. Directorate of Planning and Financial Resources, Rabat Lütkepohl H (2005) New Introduction to Multiple Time Series Analysis. Springer, Berlin, Heidelberg Granger CWJ, Newbold P (1974) Spurious regressions in econometrics. J Econom 2(2):111–120 Dickey DA, Fuller WA (1979) Distribution of the estimators for autoregressive time series with a unit root. J Am Stat Assoc 74(366a):427–431 Akaike H (1974) A new look at the statistical model identification. IEEE Trans Automat Contr 19(6):716–723 Granger CWJ (1969) Investigating Causal Relations by Econometric Models and Cross-spectral Methods. Econometrica 37(3):424–438 Sims CA (1980) Macroeconomics and Reality. Econometrica 48(1):1–48 Pfaff B (2008) VAR, SVAR and SVEC Models: Implementation Within R. J Stat Soft 27(4):1–32 Institut Royal des Etudes Stratégiques (2022) Quel système de santé au Maroc, à l'aune de la souveraineté nationale et de la généralisation de la couverture sociale ? [What health system for Morocco, in light of national sovereignty and the generalization of social coverage?] [Synthesis report]. IRES, Rabat Wagstaff A, Flores G, Hsu J, Smitz M-F, Chepynoga K, Buisman LR et al (2018) Progress on catastrophic health spending in 133 countries: A retrospective observational study. Lancet Glob Health 6(2):e169–e179 Tables Tables 1 to 4 are available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Tables.docx Cite Share Download PDF Status: Posted Version 1 posted 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-8106278","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":546591189,"identity":"41f06dd2-0de3-40a7-978f-64c18e5b94cf","order_by":0,"name":"Aazelarab BOUGHALEB","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/0lEQVRIiWNgGAWjYDACCSBmbGCQATMYKoCYmbmBKC08EC1nQFoYSdHC2MYA4eID/LObj0l+3WHHwy/dwPjh57zaaP52oJYfFdtwW3LnWJq07JlkHsk5B5gle7cdz51xmLGBsefMbdzW3Mgxk5ZsY+YxuJHAIM247VhuA1ALM2Mbbi3yEC31PPY3Eph/M845ljufkBYDoBbJj22HeQwkEtikGRtqcjcQ0mJ451iyNWPbcR6JOwfbLHuOHcjdCNRyEJ9f5G43H7z5s61aDhh0h2/8qKnLnXf+8MEHPyrweB8ImHnAFDg6DoOZB/CqB6n9gWDXEVI8CkbBKBgFIxAAAIb3Ww4rNXePAAAAAElFTkSuQmCC","orcid":"","institution":"Ibn Tofail University","correspondingAuthor":true,"prefix":"","firstName":"Aazelarab","middleName":"","lastName":"BOUGHALEB","suffix":""},{"id":546591190,"identity":"1ad12ee0-f389-43a5-917e-0cc987914e18","order_by":1,"name":"JERRY Mounir","email":"","orcid":"","institution":"Ibn Tofail University","correspondingAuthor":false,"prefix":"","firstName":"JERRY","middleName":"","lastName":"Mounir","suffix":""}],"badges":[],"createdAt":"2025-11-13 13:38:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8106278/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8106278/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96132746,"identity":"af3ef021-57e6-4ed0-8623-3e0fc118a41f","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":282759,"visible":true,"origin":"","legend":"","description":"","filename":"ArticleVARModelVF.docx","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/276467a075e62c45f321ff85.docx"},{"id":96132742,"identity":"749c1ad0-d812-4675-914e-153f88c9ba2b","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4388,"visible":true,"origin":"","legend":"","description":"","filename":"709a6386b9a442e4966a12284a64124c.json","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/e49a810e01f1f142089c3c19.json"},{"id":96132744,"identity":"425de4ae-e26d-4a65-b69c-b3d7f364b1ef","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":50545,"visible":true,"origin":"","legend":"","description":"","filename":"709a6386b9a442e4966a12284a64124c1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/2c25b9c36fd86ba82347d395.xml"},{"id":96248741,"identity":"6bc3b641-f9de-4883-94db-7254a11f7cce","added_by":"auto","created_at":"2025-11-19 07:29:05","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36116,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/34eef284c8c4ddf9688c751d.png"},{"id":96249463,"identity":"06a6d144-fa00-4ded-966f-164f9fda3bf9","added_by":"auto","created_at":"2025-11-19 07:33:38","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29676,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/f4ff6f902d6fb565453bfc26.png"},{"id":96251104,"identity":"b80af2b8-2423-4f27-93d7-94a064bca179","added_by":"auto","created_at":"2025-11-19 07:39:19","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":39756,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/5f0282f59e0bf7cfe2dc1e3d.png"},{"id":96132758,"identity":"72d12e62-925a-4007-8006-09cbbacc7091","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":21050,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/cf5197ff2b39182dc9801909.png"},{"id":96132752,"identity":"c0508220-0595-4c71-ab6f-9bfa7a2d5dfb","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15379,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/60d46607e1c8771af88d33d2.png"},{"id":96132762,"identity":"297ea4c9-1ac0-4f2a-a2f5-3d786f1fa83f","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42411,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/88df55a06521c74fcd6b2fa3.png"},{"id":96251100,"identity":"c94b28ab-59c2-4670-8caa-75f47622be49","added_by":"auto","created_at":"2025-11-19 07:39:18","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":26496,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/84f8f1b59292dfdce7db246f.png"},{"id":96132751,"identity":"7a408037-bb6e-4242-bbe3-a00ee20c16fc","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40412,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/f51207886845ea3c209ecf92.png"},{"id":96132754,"identity":"f3a9301d-b1a2-41a7-a90c-4d4391959fa9","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":10494,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/5b8087c4e559a7725d66dbb1.png"},{"id":96251111,"identity":"3f93b904-b92f-45d4-98ea-a92d680b3b89","added_by":"auto","created_at":"2025-11-19 07:39:20","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13172,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/40c9bbb79a79e698c2d97e79.png"},{"id":96132756,"identity":"344fc85c-82fd-4bb5-abad-260bbc0eb050","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13378,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/086d4b64e196bd32478c3329.png"},{"id":96251161,"identity":"a6ee3fba-d7c6-41b4-8c0b-d525ed073922","added_by":"auto","created_at":"2025-11-19 07:39:27","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9163,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/660271aa4c0379e21c882aaf.png"},{"id":96251072,"identity":"dce418e0-c3d6-4130-a9b9-00456c5fbffc","added_by":"auto","created_at":"2025-11-19 07:39:17","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7373,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/5a828c5ab8b9fb482b8cbbc8.png"},{"id":96249026,"identity":"01985247-7608-4d6b-a82f-6ca07243cd0f","added_by":"auto","created_at":"2025-11-19 07:29:57","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11078,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/c9a6ae2977df3a66792479b0.png"},{"id":96132764,"identity":"69ca5ce0-30e1-4432-8284-0c94773c0db6","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11520,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/2ef5bb01c1817460c43c2e9b.png"},{"id":96132763,"identity":"4262b81c-f2f9-497c-8cfb-0e5ee0f1a8a8","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12300,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/71693840739a4015beb1b47e.png"},{"id":96132759,"identity":"92de7707-f923-48ed-b747-1d82e116b793","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"xml","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":48289,"visible":true,"origin":"","legend":"","description":"","filename":"709a6386b9a442e4966a12284a64124c1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/2f4ba1dcecf3fb9ed5b0b3e1.xml"},{"id":96132765,"identity":"8d466c6f-a726-4413-b5c5-fe2a34cb5537","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"html","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":56154,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/6231aea4071fa6a5f2da2cb3.html"},{"id":96132739,"identity":"cb9f2e08-e3e2-4dd6-8d40-d735e79cc7a7","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":46151,"visible":true,"origin":"","legend":"\u003cp\u003eEvolution of variables in levels (2000-2022)\u003c/p\u003e\n\u003cp\u003eSource: Authors, from R Studio\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/2c16291eb887e3877a7b5de8.png"},{"id":96132740,"identity":"a95f4067-637f-42a6-bf2a-9ca8b55a00c1","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":75967,"visible":true,"origin":"","legend":"\u003cp\u003eEvolution of variables in first difference (2001-2022)\u003c/p\u003e\n\u003cp\u003eSource: Authors, from R Studio\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/71faac5b542d5577135bb389.png"},{"id":96250807,"identity":"0c2abab1-1112-4533-8461-097934cdd887","added_by":"auto","created_at":"2025-11-19 07:39:02","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":79436,"visible":true,"origin":"","legend":"\u003cp\u003eResponse to a shock on public expenditures\u003c/p\u003e\n\u003cp\u003eSource: Authors, from R Studio\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/8cb13fe49db6a671799d073f.png"},{"id":96132749,"identity":"73380d8f-ba71-499f-9d24-a5cd1e40b54d","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":58950,"visible":true,"origin":"","legend":"\u003cp\u003eGraphic of the variance decomposition (FEVD)\u003c/p\u003e\n\u003cp\u003eSource: Authors, from R Studio\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/5d510f67bf3c42965a951aba.png"},{"id":96261736,"identity":"2fcebcec-eb2e-4bcc-8f3c-223929849c38","added_by":"auto","created_at":"2025-11-19 08:00:22","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":804438,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/d9ea6612-d0f5-4fe8-a599-357280f38a77.pdf"},{"id":96132741,"identity":"3b72166e-8e31-4f9e-93d9-c4a4c4cd2c3c","added_by":"auto","created_at":"2025-11-18 03:02:32","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":108102,"visible":true,"origin":"","legend":"","description":"","filename":"Tables.docx","url":"https://assets-eu.researchsquare.com/files/rs-8106278/v1/22d6df0e2acde660da90fe41.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"The Paradox of Health Financing in Morocco: An Exploratory Analysis using a VAR Model (2000-2022)","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eThe path towards Universal Health Coverage (UHC) is a major objective of global health policies, enshrined as a central target of the Sustainable Development Goals [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. UHC is based on a dual imperative: ensuring equitable access to quality health services for the entire population and providing effective financial protection against the costs of illness [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Among the obstacles to this protection, out-of-pocket (OOP) expenditures are identified as the most inequitable mode of financing and the main driver of catastrophic health spending, capable of plunging families into poverty [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Reducing the reliance of health systems on OOP is therefore a key marker of progress towards UHC.\u003c/p\u003e\u003cp\u003eIn this global context, Morocco presents a particularly paradoxical case study. Since the beginning of the 21st century, the Kingdom has undertaken a series of ambitious reforms to expand health coverage, culminating in the royal project for the generalization of social protection launched in 2021 [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. This effort has led to a spectacular expansion of formal insurance coverage, from less than 20% of the population in the early 2000s to a goal of near-universal coverage [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. However, this progress confronts a persistent reality: despite a substantial increase in the share of public and insurance financing, the share of out-of-pocket expenditures in total health financing remains structurally high. In 2022, OOP still accounted for 43% of current health expenditures, a level well above the 15\u0026ndash;20% threshold considered a \"danger zone\" by the World Health Organization [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThis disconnect between the extension of de jure coverage and the improvement of de facto financial protection constitutes the \"Moroccan paradox.\" While the literature has well-documented the extent of this burden [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], few quantitative studies have sought to analyze the macroeconomic dynamics of this paradox over the long term. How have the interactions between public financing efforts and household spending evolved throughout the major reforms? Have public policies had a measurable and significant impact on alleviating the citizens' burden?\u003c/p\u003e\u003cp\u003eThis article aims to provide a first econometric insight into these questions. The objective is to explore the dynamic interactions between public health spending, out-of-pocket expenditures, and total health expenditure in Morocco over the period 2000\u0026ndash;2022. By mobilizing a Vector Autoregressive (VAR) model, this study seeks to characterize the nature of the dynamic relationship between these main components of financing. The goal is less to prove a definitive causality than to analyze the propagation of shocks, test for predictability, and evaluate the relative strength of mutual influences, in order to better understand the system's inertias and levers.\u003c/p\u003e\u003cp\u003eThis article is organized as follows. The next section presents the methodology and the data used. The third section presents the results of the VAR analysis, including stationarity tests, model selection, Granger causality tests, impulse response functions, and variance decomposition. Finally, the fourth section discusses the implications of these macroeconomic results and highlights the limitations of this approach, advocating for complementary micro-econometric analyses for a comprehensive understanding of the phenomenon.\u003c/p\u003e"},{"header":"METHODOLOGY AND DATA","content":"\u003cp\u003eTo analyze the dynamic interactions between the different components of health financing in Morocco, this exploratory study adopts an econometric approach based on time series. The strategy consists of modeling the relationships between key variables using a Vector Autoregressive (VAR) model, a method particularly suited for capturing dynamic interdependencies without imposing strong a priori theoretical constraints [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eData:\u003c/h2\u003e\u003cp\u003eThe data used in this study come from the World Health Organization's Global Health Expenditure Database, accessed via the World Bank's Open Data platform [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. The analysis period spans 23 years, from 2000 to 2022, providing a sufficient chronological series for a dynamic analysis. Three key variables were selected to build our model:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eOut-of-pocket expenditures (oop_pct)\u003c/b\u003e: Measured as a percentage of current health expenditures. This variable represents the direct financial burden borne by the population.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eGovernment health expenditures (gouv_pct)\u003c/b\u003e: Measured as a percentage of current health expenditures. This variable represents the collective financing effort of the state and social security schemes.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eTotal health expenditures (dts_pib_pct)\u003c/b\u003e: Measured as a percentage of Gross Domestic Product (GDP). This variable serves as a control variable to capture the overall evolution of the health sector's importance in the national economy.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eEconometric Strategy:\u003c/h3\u003e\n\u003cp\u003eThe estimation of a VAR model requires that the time series be stationary. A series is said to be stationary if its mean, variance, and autocovariance are constant over time. Most macroeconomic series exhibit trends and are therefore non-stationary in levels, which can lead to spurious regressions [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. The econometric approach therefore followed these steps:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eStationarity analysis\u003c/b\u003e: The stationarity of each series was tested using the Augmented Dickey-Fuller (ADF) test [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. The null hypothesis (H₀) of this test is the presence of a unit root, i.e., non-stationarity of the series. If the p-value of the test is greater than the 5% significance level, the null hypothesis cannot be rejected. In this case, the series is differenced (by calculating its change from one period to the next) and the test is reapplied until stationarity is achieved.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eSelection of the optimal number of lags (p)\u003c/b\u003e: A VAR(p) model expresses each variable as a linear function of its own past values and the past values of all other variables in the system, up to a lag p. The choice of p is crucial: a p that is too small may omit important dynamics, while a p that is too large consumes degrees of freedom and may overfit the model. The optimal number of lags was determined using several sequential information criteria, notably the Akaike Information Criterion (AIC) and the Schwarz Bayesian Information Criterion (SC) [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e].\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eVAR model estimation\u003c/b\u003e: Once the number of lags p was selected and the stationarity of the variables was ensured, the VAR model was estimated using the Ordinary Least Squares (OLS) method.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eAnalysis of results\u003c/b\u003e: Three main tools derived from the estimated VAR model were used for the analysis:\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eGranger causality tests\u003c/b\u003e: To determine if the past values of one variable help predict the future values of another variable [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eImpulse Response Functions (IRF)\u003c/b\u003e: To trace the path of the effect of a \"shock\" (an unexpected innovation) in one variable on the future evolution of the other variables in the system [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eForecast Error Variance Decomposition (FEVD)\u003c/b\u003e: To quantify the proportion of the forecast error variance of each variable attributable to shocks in the other variables [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eAll analyses were performed using R software (version 4.5.1) and the vars package [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e].\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cp\u003eThe application of the methodology described above allowed for the characterization of the health financing dynamics in Morocco. This section successively presents the results of the preliminary tests, model selection, and then the analyses of causality, impulse response, and variance decomposition.\u003c/p\u003e\n\u003ch3\u003eStationarity and model selection\u003c/h3\u003e\n\u003cp\u003eThe plot of the time series in levels (Fig. 1) suggests the presence of trends, particularly for public expenditures (gouv_pct) and total expenditures as a percentage of GDP (dts_pib_pct).\u003c/p\u003e\n\u003cp\u003eFormal stationarity tests confirm this visual observation. Table 1 presents the results of the Augmented Dickey-Fuller (ADF) tests. The null hypothesis of non-stationarity (presence of a unit root) is rejected at the 5% significance level for the oop_pct series (p\u0026thinsp;=\u0026thinsp;0.045), indicating that it is stationary in levels, I(0). In contrast, this hypothesis is not rejected for the gouv_pct (p\u0026thinsp;=\u0026thinsp;0.247) and dts_pib_pct (p\u0026thinsp;=\u0026thinsp;0.636) series, which are therefore considered non-stationary, I(1).\u003c/p\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003cp\u003eTo build a standard VAR model, all variables must have the same order of integration. Consequently, all series were differenced once. The ADF tests re-applied to the first-differenced series (\u0026Delta;) confirmed their stationarity (results available upon request). Figure 2 shows the evolution of these differenced series, which now oscillate around zero.\u003c/p\u003e\n\u003c/div\u003e\n\u003cp\u003eThe choice of the optimal number of lags (p) for the VAR model was guided by the information criteria presented in Table 2. The Akaike criterion (AIC), which is often favored for small samples, reaches its minimum for a lag of p\u0026thinsp;=\u0026thinsp;1. Therefore, a VAR(1) model was estimated on the first-differenced data. Diagnostic tests of the model (autocorrelation, normality, heteroscedasticity) confirmed its good specification.\u003c/p\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003eGranger causality analysis\u003c/h2\u003e\n \u003cp\u003eTable 3 presents the results of the Granger causality tests. The null hypothesis is the absence of causality from the \u0026quot;cause\u0026quot; variable to the other variables in the system. For the two relationships of interest, the p-value is well above the 5% threshold.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003cp\u003eThus, over the period studied, past variations in public spending do not provide statistically significant information to predict future variations in out-of-pocket expenditures (p\u0026thinsp;=\u0026thinsp;0.306). Likewise, the inverse causality is also not established (p\u0026thinsp;=\u0026thinsp;0.349). This absence of Granger causality suggests that the links between the annual variations of these two aggregates are complex and do not follow a simple predictive relationship, a conclusion that can be partly attributed to the low statistical power due to the limited sample size.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003eImpulse Response Analysis\u003c/h2\u003e\n \u003cp\u003eFigure 3 shows the impulse response functions (IRF) of the variation in OOP (oop_pct) and total health expenditures (dts_pib_pct) following an exogenous one-standard-deviation shock to the variation in public expenditures (gouv_pct).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003cp\u003eThe average response of the OOP variation (upper panel, black line) is negative. Immediately after the shock, the OOP variation decreases, reaching its lowest point (about \u0026minus;\u0026thinsp;0.2 percentage points) one year after the shock, before gradually returning to zero. This trajectory is consistent with economic theory: an unexpected increase in public effort tends to reduce (or slow the growth of) the household burden. However, the 95% confidence interval (red dashed lines) includes the zero line over the entire forecast horizon. This means that, although the direction of the effect is as expected, the impact is not statistically significant at the 5% level.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003eForecast Error Variance Decomposition (FEVD)\u003c/h2\u003e\n \u003cp\u003eThe variance decomposition analysis allows for the quantification of the relative contribution of each variable to the unforeseen fluctuations of the others. Table 4 details the share of the forecast error variance of the oop_pct variable explained by shocks from each variable in the system, at different time horizons.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n \u003cp\u003eThe results are unequivocal. In the short term (1-year horizon), 100% of the unexpected fluctuations in OOP are explained by its own past shocks. Even at a longer horizon of 5 years, this share remains predominant at 71.72%. Shocks to public spending (Shocks on Gov. Exp.) explain only a very marginal part of the OOP variance, reaching just 2.22% at the 5-year horizon. In contrast, shocks to total health expenditure as a percentage of GDP (Shocks on THE (%GDP)) explain a more substantial share, reaching 26.06%. These results indicate a strong inertia in the dynamics of out-of-pocket expenditures, whose variations are primarily determined by their own history, and a low sensitivity to annual budgetary policy shocks over the period studied.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThe results of the econometric analysis using a VAR model, though conducted over a period of more than two decades, present a nuanced and complex picture of health financing dynamics in Morocco. Three main findings emerge from our analyses: a strong inertia of out-of-pocket expenditures, an absence of statistically significant Granger causality, and a negative but non-significant response of OOP to public financing shocks. Far from invalidating the relevance of the issue, these results call for a deeper interpretation of the mechanisms at play and the limitations of macroeconomic analysis.\u003c/p\u003e\u003cp\u003eThe most salient result is the strong inertia of the share of out-of-pocket expenditures. The variance decomposition (Table\u0026nbsp;4) showed that more than 71% of unexpected fluctuations in OOP at a five-year horizon are explained by its own past shocks. This structural inertia suggests that the level of OOP is less sensitive to cyclical annual variations than to deep-seated factors, profoundly anchored in the architecture of the health system. This echoes the analysis of the historical trajectory of reforms, which highlighted the progressive constitution of a dual system where recourse to the private sector and direct purchase of medical goods have become firmly established practices for a large segment of the population, including the insured [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. This \"path dependence\" creates a dynamic where the behaviors of households and providers are difficult to influence through simple annual budgetary adjustments.\u003c/p\u003e\u003cp\u003eSecondly, the absence of statistically significant Granger causality between variations in public spending and those in OOP (Table\u0026nbsp;3) must be interpreted with caution. Econometrically, this means that, based on the available data, the past variations of one series do not significantly improve the prediction of the future variations of the other. Economically, this does not mean an absence of a relationship, but rather that the link is neither simple, nor direct, nor immediate. This conclusion is reinforced by the analysis of the impulse response functions (Fig.\u0026nbsp;3). Although the average response of OOP to a positive shock on public spending is negative, as expected, the effect is not statistically distinct from zero.\u003c/p\u003e\u003cp\u003eSeveral contextual factors can explain this apparent statistical disconnect:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eStatistical limitations\u003c/b\u003e: The sample size (23 annual observations) is small for a time series analysis, which considerably reduces the power of statistical tests. It is likely that a real economic relationship exists, but that the sample is not large enough to detect it with a 95% confidence level [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eAllocation of public spending\u003c/b\u003e: The increase in public health spending was not necessarily allocated to measures directly aimed at reducing out-of-pocket costs. A significant portion may have been absorbed by salary increases, the construction of new infrastructure, or the acquisition of expensive technologies, without proportionally translating into a lighter burden for households [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e].\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eCost-inflation effect in the private sector\u003c/b\u003e: It is possible that the expansion of insurance coverage created a windfall effect in a poorly regulated private sector, leading to an increase in tariffs and balance billing that would have \"absorbed\" part of the expected gains from the public effort [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e].\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThese macroeconomic results strongly underscore the need to move to a micro-level analysis. The VAR model, by construction, deals with national averages that mask very diverse realities. It cannot capture the glaring inequalities between households, nor the coping strategies, nor the impact of socio-demographic characteristics (age, chronic illness, place of residence) which are nevertheless the central determinants of the risk of catastrophic expenditure [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. The non-significance of the aggregate results therefore only reinforces the relevance of the initial research question: if the macro dynamics do not provide a clear answer, it is at the level of households and system actors that we must seek the keys to \"open the black box\" of the Moroccan paradox.\u003c/p\u003e"},{"header":"CONCLUSION","content":"\u003cp\u003eThis study aimed to explore the macroeconomic dynamics of the \"Moroccan paradox\" by analyzing the interactions between public health spending and out-of-pocket expenditures over the period 2000\u0026ndash;2022. By mobilizing a Vector Autoregressive (VAR) model, we sought to characterize the nature of this complex relationship throughout the successive reforms of the health system.\u003c/p\u003e\u003cp\u003eThe results reveal a nuanced picture. We found no statistically significant evidence of Granger causality between annual variations in public financing and those in out-of-pocket expenditures. Similarly, while the impulse response analysis suggests that a positive shock on public spending tends to reduce household expenditures, this effect is not statistically distinct from zero. The main takeaway from our model is the strong inertia in the dynamics of out-of-pocket expenditures, whose fluctuations are overwhelmingly explained by their own history, and a low sensitivity to annual budgetary policy shocks [Table\u0026nbsp;4].\u003c/p\u003e\u003cp\u003eThe main implication of these results is both methodological and substantive. From a methodological standpoint, they highlight the limits of purely macroeconomic analysis for grasping the concrete effects of health policies, particularly with short time series. From a substantive standpoint, the absence of a strong statistical relationship suggests that the link between public investment and the financial protection of households is neither direct nor automatic. This inertia points to the preponderant influence of structural factors\u0026mdash;such as the organization of the care market, weaknesses in tariff regulation, and actor behaviors\u0026mdash;which seem to hamper the full translation of public effort into an alleviation of the household burden [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThis study has limitations, the main one being the short length of the time series, which affects the power of the econometric tests. Furthermore, the use of aggregated data inherently masks the strong heterogeneities existing among households. These limitations, however, trace a clear path for future research. To truly \"open the black box\" of the Moroccan paradox, it is now imperative to change the scale of analysis. Research must move from the macroeconomic to the microeconomic, relying on household survey data. It is at this level that the real determinants of catastrophic spending risk can be robustly identified [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThis article, by providing a first quantitative characterization of the health financing dynamics in Morocco, demonstrates that while public effort in favor of health is a necessary condition, it is not sufficient on its own to guarantee the financial protection of citizens. The resolution of the paradox of high out-of-pocket expenditures lies not only in the volume of financing, but in deep structural reforms of the health system and a fine-grained understanding of microeconomic factors, which constitutes a priority research area.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eEthical Approval\u003c/h2\u003e\u003cp\u003eEthical approval was not required for this study as it is based exclusively on the analysis of publicly available, aggregated, and anonymized time-series data from the World Bank and the World Health Organization (Global Health Expenditure Database). The research did not involve human participants or access to individual-level confidential data.\u003c/p\u003e\u003ch2\u003eConflict of Interests\u003c/h2\u003e\u003cp\u003eThe authors declare that they have no competing interests to disclose.\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eThe authors declare that they received no specific funding for this research.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eA.B.: Conceptualization, Methodology, Software, Data Curation, Formal Analysis, Investigation, Writing \u0026ndash; Original Draft. M.J.: Supervision, Validation, Writing \u0026ndash; Review \u0026amp; Editing. All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets analyzed during the current study are publicly available in the World Bank Open Data repository. The data originates from the World Health Organization (WHO) Global Health Expenditure database and can be accessed via the following link: https://data.worldbank.org/indicator/SH.XPD.OOPC.CH.ZS?locations=MA.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eUnited Nations. Transforming our world: the 2030 Agenda for Sustainable Development. General Assembly (2015) Resolution A/RES/70/1\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWorld Health Organization (2010) Health systems financing: the path to universal coverage. World Health Report 2010. WHO, Geneva\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXu K, Evans DB, Carrin G, Aguilar-Rivera AM, Musgrove P, Evans T (2007) Protecting households from catastrophic health spending. Health Aff 26(4):972\u0026ndash;983\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKingdom of Morocco (2021) Framework Law 09\u0026ndash;21 on social protection. Official Gaz No 6975; April 5\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEconomic, Social and Environmental Council (2024) G\u0026eacute;n\u0026eacute;ralisation de l'AMO, bilan d'\u0026eacute;tape: Une avanc\u0026eacute;e sociale \u0026agrave; consolider, des d\u0026eacute;fis \u0026agrave; relever [Generalization of Compulsory Health Insurance, progress report: A social advance to be consolidated, challenges to be met]. Opinion No. 80/2024. CESE, Rabat\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWorld Bank Global Health Expenditure database [Internet]. World Bank Open Data. [cited June 1, 2025]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://data.worldbank.org/indicator/SH.XPD.OOPC.CH.ZS?locations=MA\u003c/span\u003e\u003cspan address=\"https://data.worldbank.org/indicator/SH.XPD.OOPC.CH.ZS?locations=MA\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMinistry of Health and Social Protection (2022) Comptes Nationaux de la Sant\u0026eacute; 2022 [National Health Accounts 2022]. Directorate of Planning and Financial Resources, Rabat\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eL\u0026uuml;tkepohl H (2005) New Introduction to Multiple Time Series Analysis. Springer, Berlin, Heidelberg\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGranger CWJ, Newbold P (1974) Spurious regressions in econometrics. J Econom 2(2):111\u0026ndash;120\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDickey DA, Fuller WA (1979) Distribution of the estimators for autoregressive time series with a unit root. J Am Stat Assoc 74(366a):427\u0026ndash;431\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAkaike H (1974) A new look at the statistical model identification. IEEE Trans Automat Contr 19(6):716\u0026ndash;723\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGranger CWJ (1969) Investigating Causal Relations by Econometric Models and Cross-spectral Methods. Econometrica 37(3):424\u0026ndash;438\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSims CA (1980) Macroeconomics and Reality. Econometrica 48(1):1\u0026ndash;48\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePfaff B (2008) VAR, SVAR and SVEC Models: Implementation Within R. J Stat Soft 27(4):1\u0026ndash;32\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eInstitut Royal des Etudes Strat\u0026eacute;giques (2022) Quel syst\u0026egrave;me de sant\u0026eacute; au Maroc, \u0026agrave; l'aune de la souverainet\u0026eacute; nationale et de la g\u0026eacute;n\u0026eacute;ralisation de la couverture sociale ? [What health system for Morocco, in light of national sovereignty and the generalization of social coverage?] [Synthesis report]. IRES, Rabat\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWagstaff A, Flores G, Hsu J, Smitz M-F, Chepynoga K, Buisman LR et al (2018) Progress on catastrophic health spending in 133 countries: A retrospective observational study. Lancet Glob Health 6(2):e169\u0026ndash;e179\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 4 are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"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":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Health financing, Out-of-Pocket expenditures (OOP), Universal Health Coverage (UHC), Financial protection, VAR model, Time series","lastPublishedDoi":"10.21203/rs.3.rs-8106278/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8106278/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDespite a spectacular expansion of its insurance coverage, Morocco faces a persistent paradox: the share of out-of-pocket (OOP) expenditures in health financing remains high, standing at around 43% in 2022. This article aims to explore the macroeconomic dynamics of this paradox by analyzing the interactions between public health spending, OOP, and total health expenditure. To do this, a Vector Autoregressive (VAR) model is applied to time series from the World Bank and WHO covering the period 2000\u0026ndash;2022. The results do not reveal statistically significant Granger causality between changes in public spending and those in OOP. The impulse response function analysis, while showing a negative response of OOP to a public spending shock, confirms the lack of statistical significance. The variance decomposition highlights a strong inertia, with over 71% of OOP fluctuations explained by their own past dynamics. These findings suggest that the relationship between public financing efforts and the alleviation of the household burden is neither direct nor automatic at the aggregate level. The study concludes on the limitations of macroeconomic analysis in capturing this complex phenomenon and underscores the imperative need to resort to micro-econometric analyses to identify the real determinants of financial risk at the household level.\u003c/p\u003e","manuscriptTitle":"The Paradox of Health Financing in Morocco: An Exploratory Analysis using a VAR Model (2000-2022)","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-18 03:02:27","doi":"10.21203/rs.3.rs-8106278/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"150e9991-f507-44ea-98b4-55eb9b8e8384","owner":[],"postedDate":"November 18th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-11-19T07:59:03+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-18 03:02:27","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8106278","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8106278","identity":"rs-8106278","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}