The Impact of Türkiye’s City Hospitals on Health Care Use and Health Outcomes

The Impact of Türkiye’s City Hospitals on Health Care Use and Health Outcomes | 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 Impact of Türkiye’s City Hospitals on Health Care Use and Health Outcomes Topal Koç Demet, Padmaja Ayyagari This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7970303/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 This study examines the impact of Türkiye’s city hospital reform on health care use and health outcomes. As part of a broader health care reform, the Turkish government established public-private partnerships to upgrade existing hospital infrastructure and build new high-capacity tertiary facilities across selected regions of the country. The first set of hospitals began operations in 2017. To estimate the effect of this reform, we employ a difference-in-differences approach, comparing changes in outcomes before and after 2017–18 in regions that opened city hospitals (treatment regions) with regions that did not open any city hospitals during the study period (control regions). We find that the reform led to a significant increase in inpatient utilization and improvements in self-reported health. Specifically, there is a 3.7 percentage point increase in the probability of using any inpatient care and 3.9 percentage point decrease in the probability of reporting bad or very bad health. Inpatient utilization outpatient utilization self-reported health public-private partnerships hospital infrastructure difference-in-differences Figures Figure 1 Figure 2 1. Introduction Middle-income countries have increasingly turned to public-private partnership (PPP) models in recent years to finance healthcare infrastructure, increase access and utilization, and improve health outcomes. Often driven by financial restrictions and capital investment deficits, these models combine private sector efficiency with public sector monitoring to improve outcomes. However, the impact of such partnerships depends crucially on several factors such as the verifiability of service quality and other institutional features, and existing empirical evidence is limited and inconclusive [ 1 ]. Increasing physical capacity may reduce allocative inefficiencies in the provision of treatment, but could also cause distortions in provider behavior, lead to an overuse of services, and worsen quality of care, especially in cases of inadequate gatekeeping and knowledge asymmetries ([ 2 ];[ 3 ]). This paper investigates a major PPP project in the health sector, known for its ambitious scope: the expansion of city hospitals (şehir hastaneleri) in Türkiye. Under long-term PPP agreements, the Turkish Ministry of Health launched a phased rollout of high-capacity tertiary care facilities across selective areas in 2017. These facilities were intended to upgrade public health infrastructure, clear regional bottlenecks in specialty treatment, and combine disparate service lines. The reform also introduced new incentive systems, most notably performance-based physician remuneration and service volume guarantees, which created concerns about over-medicalization and the mismatch of societal and private returns to healthcare services. Using nationally representative survey data, we provide the first causal estimates of the impact of Türkiye’s city hospital reform on health care utilization and health outcomes. To identify causal effects, we use a difference-in-differences approach that compares trends in utilization and health outcomes in regions that opened city hospitals between 2017 and 2018 with regions that did not open hospitals during our study period. We find significant increases in inpatient utilization and self-rated health. We also find a marginally significant increase in outpatient utilization. Our results are robust to controlling for a wide range of socioeconomic characteristics and event study regressions and placebo tests provide support for the causal interpretation. In the following section, we discuss the background of the city hospital reform in Türkiye. Section 3 reviews existing literature on public-private partnerships in health care and discusses the contribution of our study. Section 4 describes the dataset and our empirical methodology. Section 5 discusses the results and section 6 concludes. 2. Background on the City Hospital Reform in Türkiye Since the early 2000s, the Turkish government has implemented a series of healthcare reforms, with the Health Transformation Program (HTP) launched in 2003 as its cornerstone [ 4 ]. The program aimed to expand insurance coverage, strengthen primary care, and improve equity in healthcare access. While these early efforts improved health outcomes and access to outpatient care, a persistent challenge lay in Türkiye’s fragmented and insufficient hospital infrastructure, which could not meet the growing demand for specialist and inpatient care [ 5 ]. According to Health at a Glance [ 6 ], hospital bed density in Türkiye remained below the OECD average, increasing only modestly from 2.4 to 2.8 beds per 1,000 population between 2009 and 2016. Although ICU capacity was relatively high in 2021, overall bed occupancy rates were among the lowest, reflecting both excess system capacity and lower utilization during the COVID-19 pandemic. At the same time, chronic diseases have been rising among ageing populations, while regional disparities in access to secondary care have persisted [ 6 ]. In response, the government launched a second phase of reform focused on upgrading hospital infrastructure through Public-Private Partnership (PPP) . This new wave of reform aimed not only to expand capacity but also to consolidate services within modern, high-tech hospital complexes, commonly referred to as city hospitals. Project planning and feasibility assessments began in 2009. 1 Under this model, private consortia would finance, design, and maintain the facilities while the Ministry of Health retained control over medical service delivery. In exchange, the government committed to making availability payments to the firms over 25-year periods, indexed to performance metrics and often accompanied by service volume guarantees. Importantly, many contracts were denominated in foreign currency, introducing long-term fiscal risks for the public sector. In addition, the PPP hospital model adopted performance-based payments for providers, which had been introduced in 2004 as part of HTP [ 7 ]. Physicians working in city hospitals receive a fixed base salary funded from the general budget, along with performance-based bonuses which are calculated based on the volume and complexity of services delivered. Performance payments are often drawn from hospital-level revolving funds (döner sermaye). The first city hospitals opened in 2017 in Mersin, Yozgat, and Isparta, selected in part due to historical shortages in specialist care and inpatient capacity. The Turkish government continued to expand this public-private partnership (PPP) model and by 2022, over 20 city hospitals had opened, with additional projects either under construction or in the planning phase. According to the Ministry’s listing, four city hospitals opened in 2017, four more in 2018, two in 2019, five in 2020, one in 2021, and a further four by the end of 2022 ([ 8 ];[ 9 ]). Gaziantep City Hospital, a 1,875-bed facility, was completed in late 2023 and officially opened in February 2024, becoming one of the largest health complexes in southeastern Türkiye. Additional PPP projects remain under construction or tender evaluation as of 2025. Large metropolitan facilities such as Ankara Bilkent and Istanbul Başakşehir received significant investment and high-capacity healthcare infrastructure. The city hospital initiative sought to address allocative inefficiencies by expanding infrastructure in underserved areas and productive inefficiencies by centralizing services to achieve economies of scale. However, concerns persist regarding its capacity to meaningfully improve physical infrastructure and service quality. The guaranteed payments, even when demand is lower than expected, limit fiscal flexibility. Additionally, the private sector bears minimal demand risk, raising questions about incentive alignment and cost-effectiveness. Critics have also noted the potential for PPP hospitals to crowd out investments in primary care or contribute to over-medicalization ([ 10 ] ;[ 11 ];[ 12 ]). Although performance-based physician incentives included in PPP contracts can enhance productivity and reduce patient waiting times, they can also promote overuse, especially in systems with limited gatekeeping and relatively low patient co-payments ([ 13 ] ;[ 7 ] ). The sharp rise in out-of-pocket health expenditures, from 8.2 billion TL in 2009 to 220.9 billion TL in 2023 ([ 14 ];[ 15 ]), contributes to concerns about overutilization and patient welfare. Assessing the impact of these hospitals on healthcare utilization and health outcomes is therefore essential for understanding the trade-offs inherent in such large-scale infrastructure reforms. 3. Literature Review Over the past decade, governments of both developing and developed countries have increasingly implemented public-private partnerships (PPPs) in health care, with the goal of improving the delivery of health care in a cost-effective manner and ultimately improving health outcomes. Despite this growing policy relevance, the literature on PPPs in health care is comparatively underdeveloped [ 1 ]. Several studies rely on descriptive comparisons—such as contrasting PPP and public hospitals or assessing outcomes before and after PPP adoption—to evaluate their effectiveness. The findings are mixed. Evidence from Spain’s Alzira model suggests no consistent advantage over public hospitals in technical efficiency or unit costs [ 16 ] and [ 17 ]. In Lesotho, a PPP hospital network outperformed government-run facilities on clinical measures but incurred significant cost overruns [ 18 ]. In Uganda, a PPP hospital expanded outreach and facility-based deliveries but left some services unaffordable [ 19 ]. Evaluations of India’s Rashtriya Swasthya Bima Yojana (RSBY) and its successor Ayushman Bharat (PM-JAY) highlight contractual breaches and “cream skimming,” with private providers favoring low-cost, high-margin services, problems attributed to weak regulation and monitoring [ 20 ]. Overall, the literature shows no clear consensus on the impact of PPPs in health care, reflecting wide variation in how they are designed, implemented, and assessed [ 21 ]. Moreover, most descriptive studies cannot rule out alternative explanations, such as secular trends, for observed effects. Fabre and Straub [ 1 ] focus on studies evaluating PPPs in the health sector that employ causal inference methods such as difference-in-differences. Overall, the evidence is mixed, much like the findings from descriptive studies. For example, Mohanan et al. [ 22 ] found that India’s Chiranjeevi Yojana program had no impact on institutional delivery rates or maternal outcomes, possibly because the quality of participating facilities was perceived as low or because private providers were still charging fees for facility delivery. In Afghanistan, Engineer et al. [ 23 ] evaluated a pay-for-performance program designed to improve maternal and child health services and found no overall effect, partly because health workers did not fully understand the payment incentives. However, they did observe increases in patient interaction time and quality of care. In Congo, Huillery and Seban [ 24 ] found that a PPP incentivizing providers to increase utilization paradoxically reduced both utilization and provider revenue, driven not by lower effort but by patients’ perception of reduced care quality. By contrast, Soeters et al. [ 25 ] evaluated another performance-based program in Congo, which rewarded both service utilization and care quality, and found improvements in financial access, facility revenues, and quality of care. Cristia et al. [ 26 ] report that a PPP involving mobile medical teams in Guatemala substantially increased children’s immunization rates and shifted prenatal care provision from traditional midwives to physicians and nurses. In Cameroon, De Walque et al. [ 27 ] found that a facility-based program raised utilization for some services, such as vaccinations and family planning, but not for others such as antenatal care and facility-based deliveries. In Kenya, Obare et al. [ 28 ] showed that a reproductive health voucher program increased institutional deliveries and skilled birth attendance, though it had no effect on antenatal care. Synthesizing these findings, Fabre and Straub [ 1 ] conclude that PPPs are more effective when incentive payments are tied to both service volume and quality, and when demand- and supply-side policies are combined to improve utilization and outcomes. Given the mixed findings, further research is needed to better understand the conditions under which PPPs improve access and health outcomes. Our study contributes to this literature by providing the first evaluation of Türkiye’s city hospital reform, one of the largest PPP initiatives undertaken in the health sector. We employ a difference-in-differences approach to identify causal effects, thereby contributing new evidence on the impact of this reform and more generally, on the impact of PPPs in the health care sector. 4. Data and Empirical Methodology We use repeated cross-sectional data from the 2008, 2010, 2014, 2016, 2019, and 2022 waves of the Türkiye Health Survey, a nationally representative survey administered by the Turkish Statistical Institute. The THS collects comprehensive data on individuals' healthcare use, chronic illnesses, perceived health, health practices, and sociodemographic background. The sample design guarantees comparability across survey waves and representativeness at the NUTS-2 regional level. 2 Our dependent variables include binary indicators for any inpatient or outpatient service use in the past year, and an indicator for self-reported bad or very bad health (very good, good, and fair health form the reference category). Independent variables include binary indicators for female, age categories (25–44 years, 45–54 years, 55–64 years, and \(\:\ge\:\) 65 years, with < 25 years as the reference category), married status (the reference category includes widowed, divorced, and never married), income categories (middle and high with poor as the reference group), missing income, and education level (secondary and tertiary with primary as the reference group). Household income was reported in categorical brackets that varied across survey years (5 categories in 2014–2016, 20 categories in 2019–2022). To ensure comparability across survey waves, we collapsed income into three groups that were consistently defined across waves: low income (0–1,200 TL or approximately 450–500 USD), middle income (1,200–2,500 TL or approximately 500–850 USD), and high income (≥ 2,500 TL or approximately 850 USD+). Missing responses were coded as a separate dummy variable. We categorize individuals as belonging to the treatment or control group based on their NUTS-2 region of residence, which is the most granular geographic identifier available in the survey. The control group includes individuals living in areas without any city hospital openings throughout our study period (2008–2022), and the treatment group includes individuals living in NUTS-2 areas where city hospitals were opened in 2017 or 2018. We exclude NUTS-2 regions where city hospitals were opened after 2018 to allow a sufficient post period necessary for the difference-in-differences analysis (described below) and to avoid bias in two-way fixed effects models due to the staggered rollout and heterogeneity in treatment effects [ 29 ]. Figure 1 presents the distribution of treatment, control, and excluded NUTS-2 regions across Türkiye. Treatment regions are depicted in red, control regions in light pink, and excluded regions are in gray. After excluding observations with missing or inconsistent values and in regions where city hospitals opened after 2018, our analysis sample includes 115,673 observations, with 27,460 in treatment regions and the remaining in control regions. Table 1 presents summary statistics for health care use, health outcomes, and sociodemographic characteristics for our analysis sample, and separately for the treatment and control groups. Nearly 54% of the sample is female, 13% are aged 65 or older, and 13.7% have completed tertiary education. About 10% report any inpatient use in the past 12 months, while 63% report using outpatient care. The sample is relatively healthy, with fewer than 10% reporting bad or very bad health. The treatment and control groups are similar across most observable characteristics, although the control group has slightly higher income and educational attainment. Importantly, our empirical strategy is a difference-in-differences design, which does not require balance between the two groups, but rather relies on the assumption of parallel trends. Table 1 Summary Statistics (1) (2) (3) Full Sample Control Group Treatment Group Any Inpatient Use 0.101 0.101 0.102 (0.302) (0.301) (0.303) Any Outpatient Use 0.633 0.632 0.636 (0.482) (0.482) (0.481) Bad or Very Bad Health 0.096 0.089 0.118 (0.294) (0.284) (0.323) Female 0.538 0.538 0.538 (0.499) (0.499) (0.499) Married 0.685 0.684 0.691 (0.464) (0.465) (0.462) Age 25–44 0.391 0.393 0.385 (0.488) (0.488) (0.487) Age 45–54 0.175 0.176 0.173 (0.380) (0.380) (0.378) Age 55–64 0.133 0.132 0.137 (0.340) (0.339) (0.344) Age > = 65 0.130 0.130 0.131 (0.337) (0.337) (0.337) Middle Income 0.387 0.388 0.384 (0.487) (0.487) (0.486) High Income 0.220 0.230 0.186 (0.414) (0.421) (0.389) Missing Income 0.259 0.259 0.260 (0.438) (0.438) (0.439) Secondary Education 0.142 0.144 0.135 (0.349) (0.351) (0.342) Tertiary Education 0.137 0.140 0.125 (0.344) (0.347) (0.331) Observations 115,673 88,213 27,460 Source: Türkiye Health Survey Data (2008–2022). Notes: Table presents means and standard deviations are in parentheses. To identify the causal effect of the reform on health care use and health outcomes, we employ the following difference-in-differences regression model: \(\:{Y}_{it}={\beta\:}_{0}+{\beta\:}_{1}Treatmen{t}_{i}\times\:Pos{t}_{t}+{\beta\:}_{2}Treatmen{t}_{i}+{\beta\:}_{3}Pos{t}_{t}+{\beta\:}_{4}{X}_{it}+{\epsilon\:}_{it}\) (Eq. 1) Where, \(\:{Y}_{it}\) represents inpatient care use, outpatient care use, or health outcomes for person \(\:i\:\) in survey year \(\:t\) . \(\:Treatmen{t}_{i}\) is one if person \(\:i\) resides in a region with a city hospital opening and zero if they reside in a region without any city hospital openings during the study period. \(\:Pos{t}_{t}\) is one for observations in 2019 and 2022 and zero for observations in 2008, 2010, 2014, or 2016. \(\:{X}_{it}\) is a vector of sociodemographic characteristics (gender, age, marital status, income, and education) described above. In addition, \(\:{X}_{it}\) includes region-specific linear trends which account for unobserved, time-varying factors that may be correlated with differential trends in health care use and health outcomes. Following the discussion in Abadie et al. [ 30 ], standard errors are clustered at the treatment assignment level which is the NUTS-2 region level in our application. Since there are only 20 NUTS-2 regions in our sample, we use a wild cluster bootstrap with Webb weights to adjust for possible downward bias in standard errors originating from a small number of clusters [ 31 ]. The coefficient of interest is \(\:{\beta\:}_{1}\) , which identifies the causal effect of the reform on outcomes under the parallel trends assumption. With the inclusion of region-specific linear trends, identification relies on the assumption that, in the absence of the reform, any differences between treated and control groups follow parallel nonlinear trends. This is less restrictive than the standard parallel trends assumption, as it allows for differential linear trends across regions. To assess the validity of this assumption, we use an event study regression model which replaces the \(\:Pos{t}_{t}\) variable in Eq. (1) with a full set of year fixed effects. \(\:{Y}_{it}={\beta\:}_{0}+\sum\:_{t}{\beta\:}_{1t}Treatmen{t}_{i}\times\:Yea{r}_{t}+{\mu\:}_{g}+{\sum\:_{t}{\beta\:}_{2t}Year}_{t}+{\beta\:}_{3}{X}_{it}+{\epsilon\:}_{it}\) (Eq. 2) Where, \(\:Yea{r}_{t}\) represents year fixed effects (with 2016 as the reference year), \(\:{\mu\:}_{g}\) represents region fixed effects, \(\:\:\) and all other terms are defined as above, except that \(\:{X}_{it}\) does not include region specific linear trends. A lack of differential trends between treatment and control regions in the pre period (2008–2016) would provide support for the identifying assumption. As a further check of our identification strategy, we conduct a placebo test by splitting the control group into two arbitrary subsets, designating one as a “placebo treatment” group and the other as a “placebo control” group and re-estimating the difference-in-differences model using only control observations. Since none of the observations in the control group were exposed to the opening of a city hospital during the study period, the difference-in-differences estimate should be statistically indistinguishable from zero if the identifying assumption is satisfied. On the other hand, a significant effect would be an indication of unobserved differential trends across regions which may explain the main estimates. 5. Results Table 2 presents the difference-in-differences estimates of the impact of city hospital openings on inpatient and outpatient healthcare use and on self-reported health. Odd numbered columns present results from a specification that does not include any covariates while even numbered columns present results from a specification that includes sociodemographic characteristics and region-specific linear trends. Wild cluster bootstrap p-values are presented in brackets and are used for inference. For reference, we also present conventional robust standard errors clustered at the NUTS-2 region level. Including covariates improves the precision of the difference-in-differences estimates, and therefore, we focus on the specification with covariates as our preferred specification. We find that city hospital openings significantly increased inpatient care use and improved self-reported health. Specifically, there is 3.7 percentage point increase in the probability of using any inpatient care in response to the reform (Column 2). This represents a 36.6% increase relative to the sample mean. The use of outpatient care also increased by 7.3 percentage points, but this estimate is only significant at the 10% level (Column 4). The program reduced reports of bad or very bad health by 3.9 percentage points (Column 6), which represents a 40.6% relative decrease. Table 2 Effect of City Hospitals on Health Care Use and Health Outcomes (1) (2) (3) (4) (5) (6) Any Inpatient Use Any Inpatient Use Any Outpatient Use Any Outpatient Use Bad or Very Bad Health Bad or Very Bad Health Treatment X Post 0.014** 0.037*** 0.011 0.073* -0.017* -0.039*** (0.005) (0.006) (0.028) (0.041) (0.009) (0.010) [0.009] [0.000] [0.732] [0.287] [0.169] [0.010] Post 0.000 -0.023*** 0.041*** -0.074*** -0.010** 0.013 (0.003) (0.007) (0.010) (0.015) (0.004) (0.009) [0.957] [0.003] [0.022] [0.001] [0.060] [0.175] Treatment -0.003 0.014** 0.001 0.060*** 0.034*** -0.000 (0.006) (0.006) (0.021) (0.017) (0.012) (0.011) [0.734] [0.081] [0.977] [0.003] [0.021] [0.987] Covariates X X X N 115,618 115,618 115,670 115,670 115,651 115,651 Adjusted R2 0.000 0.022 0.002 0.046 0.002 0.121 Source: Türkiye Health Survey Data (2008–2022). Notes: Covariates include gender, age, marital status, income, and educational attainment. Robust standard errors clustered at the NUTS2 level are in parentheses. Wild bootstrap p-values based on 9,999 replications are in brackets. *p < 0.10 **p < 0.05 ***p < 0.001 Next, we examine the validity of the identifying assumption using the event study regression. Figure 2 presents the coefficients on the interaction between the treatment dummy and year fixed effects, together with 95% confidence intervals. The interaction for 2016 serves as the reference category. For inpatient use, the pre-trend is relatively flat, with a clear increase in utilization in the 2019 and 2022 waves relative to 2016, consistent with a treatment effect emerging only after the reform. For outpatient use, the pre-trend is similarly flat; there is little change in 2019, followed by a modest uptick that emerges in 2022. For self-reported health, the pre-trend shows an upward trend that is not statistically significant, followed by a downward trend in the post-year period. Overall, the results provide strong support for the assumption of parallel trends for inpatient and outpatient use. In the case of self-reported health, while the graph suggests that the difference between treatment and control regions is increasing over the pre-period, the lack of statistical significance implies that we cannot rule out parallel trends. Moreover, the inclusion of region-specific linear trends in our preferred specification suggests that the identifying assumption of parallel non-linear trends is likely to be satisfied. As a further check on the identifying assumptions, Table 3 presents results from the placebo regressions. For this analysis, the sample is restricted to individuals residing in control regions where no city hospital was constructed during our study period. Since these individuals do not have access to a city hospital, we do not expect the reform to affect their health care use or health outcomes. Thus, any statistically significant effects would indicate spurious treatment effects and a potential violation of the identifying assumptions. Reassuringly, none of the placebo difference-in-differences estimates for inpatient use, outpatient use, or self-reported health are statistically significant. Moreover, the placebo estimates have the opposite sign of our main difference-in-differences estimates, further suggesting that the main results are not driven by unobserved factors contributing to differential trends between treatment and control regions. Together with the evidence from the event study regressions, these findings provide strong support for the validity of the identifying assumptions. Table 3 Placebo Tests (1) (2) (3) Any Inpatient Use Any Outpatient Use Bad or Very Bad Health Placebo Treatment X Post -0.003 -0.001 0.028 (0.011) (0.041) (0.020) [0.786] [0.988] [0.242] Post -0.040*** -0.045** -0.041*** (0.007) (0.018) (0.007) [0.005] [0.008] [0.005] Placebo Treatment 0.022* 0.049* 0.035* (0.010) (0.027) (0.020) [0.080] [0.108] [0.129] N 88,178 88,210 88,198 Adjusted R2 0.022 0.046 0.110 Source: Türkiye Health Survey Data (2008–2022). Notes: All regressions include gender, age, marital status, income, and educational attainment as covariates. Robust standard errors clustered at the NUTS2 level are in parentheses. Wild bootstrap p-values based on 9,999 replications are in brackets. *p < 0.10 **p < 0.05 ***p < 0.001 6. Conclusion and Discussion Using nationally representative data and a difference-in-differences methodology, we find that Türkiye’s city hospital reform significantly increased inpatient utilization by 3.7 percentage points, which represents a sizeable relative increase of 36.6% given the low rates of inpatient utilization in this population. There is some evidence of an increase in outpatient utilization, but the effect is only significant at the 10% level. We also find a 3.9 percentage point decrease in self-reports of bad or very bad health, which represents a 40.6% relative decline. The increase in utilization combined with health improvements suggests that the reform meaningfully expanded access to hospital services and mitigates concerns about physician induced demand or the overuse of care. Our findings are broadly consistent with studies showing that PPPs can improve service delivery when performance-based incentives are linked to utilization and quality outcomes [ 25 ], [ 27 ]. However, the results differ from evidence in countries such as India and Afghanistan, where incentive programs had limited effects due to weak implementation or low perceived quality [ 22 ], [ 23 ]. The Turkish experience shows that combining infrastructure expansion with performance-linked physician incentives can meaningfully enhance access and health outcomes, particularly when capacity constraints are more severe. Our study offers important insights into how infrastructure policy and provider incentives interact in shaping health capital. The findings suggest that building tertiary care facilities, particularly in underdeveloped areas, can enhance service utilization when paired with performance-based incentives to increase provider availability and effort. However, due to data constraints, our analysis is limited to a select set of utilization and health measures and covers only two years of the post-reform period. To fully assess the impact of this reform, future research should examine longer-term changes in access and health, incorporate more detailed measures of utilization and health outcomes, and explore the effects of expanding city hospitals into other regions, which may face different challenges or experience varying impacts. Declarations Ethics approval and consent to participate This study uses de-identified secondary microdata from the Turkish Statistical Institute (TURKSTAT, TÜİK) Türkiye Health Survey (THS) 2008–2022, accessed under a formal data use permission. The data contain no personally identifiable information, and the analyses involved no direct contact with human participants or experimental interventions. All procedures were performed in accordance with the Turkish Statistical Law No. 5429 and TURKSTAT’s confidentiality and data security standards. Therefore, the study is exempt from institutional ethics review, and no additional ethical approval was required. Informed consent was obtained by TURKSTAT during the data collection process. Consent for publication Not applicable. The study uses secondary anonymized data that do not contain identifiable individual information. Competing interests The authors declare that they have no financial, non-financial, business, or family competing interests. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. However, the first author acknowledges support from the Scientific and Technological Research Council of Türkiye (TÜBİTAK) 2219 International Postdoctoral Research Fellowship Program. Author Contribution DTK (Kirklareli University, University of South Florida): Conceptualization, data curation, formal analysis, methodology, writing original draft preparation. PA (University of South Florida): Supervision, validation, writing, review & editing. All authors read and approved the final manuscript. Acknowledgement The author thanks the Scientific and Technological Research Council of Türkiye (TÜBİTAK). Data Availability The data supporting the findings of this study are taken from the Türkiye Health Survey (THS) conducted by the Turkish Statistical Institute (TURKSTAT). Due to data confidentiality policies, these microdata cannot be shared publicly by the authors. Researchers may request access through the official TURKSTAT microdata application process:[https://www.tuik.gov.tr/Kurumsal/Mikro\_Veri](https:/www.tuik.gov.tr/Kurumsal/Mikro_Veri) . References Fabre A, Straub S. The Impact of Public–Private Partnerships (PPPs) in Infrastructure, Health, and Education. J Econ Lit. 2023;61:655–715. https://doi.org/10.1257/JEL.20211607 . Mcguire TG. Handbook of Health Economics. In: Culyer AJ, Newhouse JP, editors. Handbook of Health Economics. 1A edition. Amsterdam: Elsevier Science B.V.; 2000. pp. 461–536. https://doi.org/10.1016/S1574-0064(00)80168-7 . Propper C, Reenen J, Van. Can Pay Regulation Kill? Panel Data Evidence on the Effect of Labor Markets on Hospital Performance. J Polit Econ. 2010;118:189–408. https://doi.org/10.1086/653137 . Tatar M, Mollahaliloǧlu S, Sahin B, Aydin S, Maresso A, Hernández-Quevedo C. Turkey Health system review. Health Syst Transit. 2011;13:1–186, xiii. Turkish Ministry of Health, (Ed. Prof. Dr. Recep Akdağ). Turkey Health Transformation Program (2003–2010): Evaluation Report. Ankara; 2011. OECD. Health at a Glance 2023: OECD Indicators. Paris: OECD Publishing; 2023. https://doi.org/10.1787/7a7afb35-en . Aktas P, Hammond J, Richardson L. Pay-for-performance in healthcare provision: the role of discretion in policy implementation in Turkey. Int J Public Sect Manag. 2023;36:530–45. https://doi.org/10.1108/IJPSM-12-2022-0282 . Ministry of Health of Turkey Directorate General of Health Investments. Ten city hospitals opened in three years. 2020. https://sygm.saglik.gov.tr/TR,58066/3-yilda-10-sehir-hastanesi-hizmete-acildi.html . Accessed 2 Aug 2025. Ali D, Benli RSD, Platform. The modern face of Turkey: City hospitals. 2024. https://sdplatform.com/turkiyenin-modern-yuzu-sehir-hastaneleri/#:~:text=Adana%2C Elazığ%2C Eskişehir%2C Isparta,adet şehir hastanesi işletmeye açılmıştır. Accessed 30 Jul 2025. Ari HO, Şahi̇n B. Evaluation of health personnel’s views on city hospitals (CH’s) as a public-private partnership practice in the Turkish health system. Int J Healthc Manag. 2024;17:756–65. https://doi.org/10.1080/20479700.2023.2235790 . Demirag I, Fırtın CE, Tekin Bilbil E. Managing expectations with emotional accountability: making City Hospitals accountable during the COVID-19 pandemic in Turkey. J Public Budg Acc Financ Manag. 2020;32:889–901. https://doi.org/10.1108/JPBAFM-07-2020-0097 . Özzeybek Taş M, Tengilimoglu D, Atilla EA, A Research For Determining The Opinions On City Hospitals Example Of Ankara. The Capital City. A Res Determ Opin city Hosp Ex Ankara. Cap city. 2019;4:140–59. https://doi.org/10.31201/IJHMT.614614 . Lee DW, Jang J, Choi DW, Jang SI, Park EC. The effect of shifting medical coverage from National Health Insurance to Medical Aid type I and type II on health care utilization and out-of-pocket spending in South Korea. BMC Health Serv Res. 2020;20:1–10. https://doi.org/10.1186/s12913-020-05778-2 . Turkish Statistical Institute (TURKSTAT). 2023 Health Expenditure Statistics of Türkiye. Turkish Statistical Institute News. 2024. https://data.tuik.gov.tr/Bulten/Index?p=Saglik-Harcamalari-Istatistikleri-2023-53561 . Accessed 2 Aug 2025. Turkish Statistical Institute (TURKSTAT). 2009–2012 Health Expenditure Statistics of Türkiye, Turkish Statistical Institute News. 2013. https://data.tuik.gov.tr/Bulten/Index?p=Health-Expenditure-Statistics-2009-2012-15871 . Accessed 2 Aug 2025. Comendeiro-Maaløe M, Ridao-López M, Gorgemans S, Bernal-Delgado E. Public-private partnerships in the Spanish National Health System: The reversion of the Alzira model. Health Policy (New York). 2019;123:408–11. https://doi.org/10.1016/j.healthpol.2019.01.012 . Caballer-Tarazona M, Vivas-Consuelo D. A cost and performance comparison of Public Private Partnership and public hospitals in Spain. Health Econ Rev. 2016;6:1–7. https://doi.org/10.1186/s13561-016-0095-5 . McIntosh N, Grabowski A, Jack B, Nkabane-Nkholongo EL, Vian T. A public-private partnership improves clinical performance in a hospital network in Lesotho. Health Aff. 2015;34:954–62. https://doi.org/10.1377/hlthaff.2014.0945 . Asasira J, Ahimbisibwe F. Public-Private Partnership in Health Care and Its Impact on Health Outcomes: Evidence from Ruharo Mission Hospital in Uganda. Int J Soc Sci Stud. 2018;6:79. https://doi.org/10.11114/ijsss.v6i12.3911 . Khetrapal S, Acharya A, Mills A. Assessment of the public-private-partnerships model of a national health insurance scheme in India. Soc Sci Med. 2019;243. https://doi.org/10.1016/j.socscimed.2019.112634 . Rodrigues NJP. Public–Private Partnerships Model Applied to Hospitals—A. Crit Rev Healthc. 2023;11. https://doi.org/10.3390/healthcare11121723 . Mohanan M, Bauhoff S, La Forgia G, Babiarz KS, Singh K, Miller G. Effect of Chiranjeevi Yojana on institutional deliveries and neonatal and maternal outcomes in Gujarat, India: a difference-in-differences analysis. Bull World Heal Organ. 2014;92:187–94. https://doi.org/10.2471/BLT.13.124644 . Engineer CY, Dale E, Agarwal A, Agarwal A, Alonge O, Edward A, et al. Effectiveness of a pay-for-performance intervention to improve maternal and child health services in Afghanistan: a cluster-randomized trial. Int J Epidemiol. 2016;45:451–9. https://doi.org/10.1093/IJE/DYV362 . Huillery E, Seban J. Financial incentives, efforts, and performances in the health sector: Experimental evidence from the Democratic Republic of Congo. Econ Dev Cult Change. 2021;69. https://doi.org/https://doi.org/10.1086/703235 . Soeters R, Peerenboom PB, Mushagalusa P, Kimanuka C. Performance-Based Financing Experiment Improved Health Care In The Democratic Republic Of Congo. Health Aff. 2011;30. https://doi.org/10.1377/hlthaff.2009.0019 . Cristia J, Evans WN, Kim B. Improving the Health Coverage of the Rural Poor: Does Contracting-Out Mobile Medical Teams Work? J Dev Stud. 2015;51:247–61. https://doi.org/10.1080/00220388.2014.976617 . De Walque D, Robyn PJ, Saidou H, Sorgho G, Steenland M. Looking into the performance-based financing black box: evidence from an impact evaluation in the health sector in Cameroon. Health Policy Plan. 2021;36:835. https://doi.org/10.1093/heapol/czab002 . Obare F, Warren C, Njuki R, Abuya T, Sunday J, Askew I, et al. Community-level impact of the reproductive health vouchers programme on service utilization in Kenya. Health Policy Plan. 2013;28:165–75. https://doi.org/10.1093/HEAPOL/CZS033 . Goodman-Bacon A. Difference-in-differences with variation in treatment timing. J Econom. 2021;225:254–77. https://doi.org/10.1016/J.JECONOM.2021.03.014 . Abadie A, Athey S, Imbens GW, Wooldridge JM. When Should You Adjust Standard Errors for Clustering? Q J Econ. 2022;138:1–35. https://doi.org/10.1093/QJE/QJAC038 . Colin Cameron A, Miller DL. A Practitioner’s Guide to Cluster-Robust Inference. J Hum Resour. 2015;50:317–72. https://doi.org/10.3368/JHR.50.2.317 . Republic of Turkey Official Gazette. Law No. 6428: Law on the construction of facilities through public–private partnerships in the field of healthcare services and amendments to certain laws and statutory decrees. Ankara; 2013. Footnotes These efforts were formalized through the passage of Law No. 6428 in 2013, enabling PPPs in health infrastructure [32]. NUTS-2 regions refer to the second level administrative regions according to Nomenclature of Territorial Units for Statistics. There are 26 NUTS-2 regions in Türkiye. Additional Declarations No competing interests reported. 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-7970303","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":545696546,"identity":"60d94b4d-fc94-43cb-9e74-0f1f24494867","order_by":0,"name":"Topal Koç Demet","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABAElEQVRIiWNgGAWjYDCCAzAGOxuQqEiQAbElCGo5kAAkmEFaziTwQLUYEKmFsY0ILXzHuxMff/xhk8ffzJb44Oe8NB6DA8wHb/Mw/MnHpUXyzNnNBgcS0oolDrMdNuzdlgPUwpZszcNgYNmAQ4vBjdxtEgcSDic2HGZvk2bcVgHUwmMmDdSC02UG999u/wHSMh+sZQ5IC/83/Fpu8G5jAGnZcJjtmDRjA8hhPGx4tUieyd0scSYtrdjwMFuyYc+xNB7Jw2zGlnMMjHFq4Tt+duOHChubPLnjbYYPftQky/Edb354402FHJ6IgYAEBJMZ7GBCGlC0jIJRMApGwShAAwAhW1bTgPLFzQAAAABJRU5ErkJggg==","orcid":"","institution":"Kırklareli University","correspondingAuthor":true,"prefix":"","firstName":"Topal","middleName":"Koç","lastName":"Demet","suffix":""},{"id":545696547,"identity":"8abdfbd3-e123-4b94-b13b-9da9bf330e99","order_by":1,"name":"Padmaja Ayyagari","email":"","orcid":"","institution":"University of South Florida","correspondingAuthor":false,"prefix":"","firstName":"Padmaja","middleName":"","lastName":"Ayyagari","suffix":""}],"badges":[],"createdAt":"2025-10-28 16:49:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7970303/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7970303/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96079632,"identity":"14c3ba29-cffa-43da-9dd5-956a7a1a9a35","added_by":"auto","created_at":"2025-11-17 11:19:11","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":291676,"visible":true,"origin":"","legend":"","description":"","filename":"TheImpactofTrkiyesCityHospitalsonHealthCareUseandHealthOutcomes.docx","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/c50324cbe1deb605013a951f.docx"},{"id":96079630,"identity":"896fc940-5ebc-46be-8072-0ac0bc2def9e","added_by":"auto","created_at":"2025-11-17 11:19:11","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4291,"visible":true,"origin":"","legend":"","description":"","filename":"558665c289304fa99a3585904026e7af.json","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/eb1d4f50ce7e89456654701e.json"},{"id":96079631,"identity":"ebdddd43-6cb3-45bf-ac97-48b1592bb46f","added_by":"auto","created_at":"2025-11-17 11:19:11","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":104853,"visible":true,"origin":"","legend":"","description":"","filename":"558665c289304fa99a3585904026e7af1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/f2bbf6e8b1aef669325fc053.xml"},{"id":96079629,"identity":"654c3337-fa6a-4c98-9f4d-5c4c4a53add6","added_by":"auto","created_at":"2025-11-17 11:19:11","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":31231,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/7d4da5193ca73a0ed7769817.png"},{"id":96079633,"identity":"cc20c5d1-b651-466b-b317-0b911faf1638","added_by":"auto","created_at":"2025-11-17 11:19:11","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":16598,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/ea8c0545887923f9f39e1bab.png"},{"id":96079638,"identity":"abe8bb40-8c3f-4d14-be26-af64d7afe123","added_by":"auto","created_at":"2025-11-17 11:19:19","extension":"xml","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":105029,"visible":true,"origin":"","legend":"","description":"","filename":"558665c289304fa99a3585904026e7af1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/fd146b560c04f2d36ebf2b5c.xml"},{"id":96079634,"identity":"a8cdfc64-cbfb-4a48-b61f-b2bc2efafd3e","added_by":"auto","created_at":"2025-11-17 11:19:11","extension":"html","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":112900,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/4e3e4f40ea1d9ea906404713.html"},{"id":96079635,"identity":"24b350bb-a584-4e99-aad0-d8312814a390","added_by":"auto","created_at":"2025-11-17 11:19:13","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":121423,"visible":true,"origin":"","legend":"\u003cp\u003eGeographic distribution of treatment and control regions in Türkiye\u003c/p\u003e\n\u003cp\u003eSource: Türkiye Health Survey Data (2008-2022). Notes: Treatment regions are shaded in red, excluded regions in gray, and control regions in light pink on the Türkiye NUTS-2 regions map.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/5baeb199b14b0e3b6e2e155f.png"},{"id":96079637,"identity":"e1311a51-027c-4c11-a418-00b8c55b13f3","added_by":"auto","created_at":"2025-11-17 11:19:16","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":69962,"visible":true,"origin":"","legend":"\u003cp\u003eEvent Study Regressions\u003c/p\u003e\n\u003cp\u003eSource: Türkiye Health Survey Data (2008-2022).\u003c/p\u003e\n\u003cp\u003eNotes: Graph presents interaction coefficients and 95% confidence intervals based on a wild bootstrap method from an event study regression.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/3e672ded37822d7c4a975e81.png"},{"id":96256116,"identity":"4854e0a9-6c1e-4732-be93-4264d074348f","added_by":"auto","created_at":"2025-11-19 07:49:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":778920,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7970303/v1/11063441-3894-455a-8961-645ba93c20b8.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"The Impact of Türkiye’s City Hospitals on Health Care Use and Health Outcomes","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eMiddle-income countries have increasingly turned to public-private partnership (PPP) models in recent years to finance healthcare infrastructure, increase access and utilization, and improve health outcomes. Often driven by financial restrictions and capital investment deficits, these models combine private sector efficiency with public sector monitoring to improve outcomes. However, the impact of such partnerships depends crucially on several factors such as the verifiability of service quality and other institutional features, and existing empirical evidence is limited and inconclusive [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Increasing physical capacity may reduce allocative inefficiencies in the provision of treatment, but could also cause distortions in provider behavior, lead to an overuse of services, and worsen quality of care, especially in cases of inadequate gatekeeping and knowledge asymmetries ([\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e];[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]).\u003c/p\u003e\u003cp\u003eThis paper investigates a major PPP project in the health sector, known for its ambitious scope: the expansion of city hospitals (şehir hastaneleri) in T\u0026uuml;rkiye. Under long-term PPP agreements, the Turkish Ministry of Health launched a phased rollout of high-capacity tertiary care facilities across selective areas in 2017. These facilities were intended to upgrade public health infrastructure, clear regional bottlenecks in specialty treatment, and combine disparate service lines. The reform also introduced new incentive systems, most notably performance-based physician remuneration and service volume guarantees, which created concerns about over-medicalization and the mismatch of societal and private returns to healthcare services.\u003c/p\u003e\u003cp\u003eUsing nationally representative survey data, we provide the first causal estimates of the impact of T\u0026uuml;rkiye\u0026rsquo;s city hospital reform on health care utilization and health outcomes. To identify causal effects, we use a difference-in-differences approach that compares trends in utilization and health outcomes in regions that opened city hospitals between 2017 and 2018 with regions that did not open hospitals during our study period. We find significant increases in inpatient utilization and self-rated health. We also find a marginally significant increase in outpatient utilization. Our results are robust to controlling for a wide range of socioeconomic characteristics and event study regressions and placebo tests provide support for the causal interpretation.\u003c/p\u003e\u003cp\u003eIn the following section, we discuss the background of the city hospital reform in T\u0026uuml;rkiye. Section 3 reviews existing literature on public-private partnerships in health care and discusses the contribution of our study. Section 4 describes the dataset and our empirical methodology. Section 5 discusses the results and section 6 concludes.\u003c/p\u003e"},{"header":"2. Background on the City Hospital Reform in Türkiye","content":"\u003cp\u003eSince the early 2000s, the Turkish government has implemented a series of healthcare reforms, with the Health Transformation Program (HTP) launched in 2003 as its cornerstone [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. The program aimed to expand insurance coverage, strengthen primary care, and improve equity in healthcare access. While these early efforts improved health outcomes and access to outpatient care, a persistent challenge lay in T\u0026uuml;rkiye\u0026rsquo;s fragmented and insufficient hospital infrastructure, which could not meet the growing demand for specialist and inpatient care [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAccording to Health at a Glance [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e], hospital bed density in T\u0026uuml;rkiye remained below the OECD average, increasing only modestly from 2.4 to 2.8 beds per 1,000 population between 2009 and 2016. Although ICU capacity was relatively high in 2021, overall bed occupancy rates were among the lowest, reflecting both excess system capacity and lower utilization during the COVID-19 pandemic. At the same time, chronic diseases have been rising among ageing populations, while regional disparities in access to secondary care have persisted [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. In response, the government launched a second phase of reform focused on upgrading hospital infrastructure through \u003cem\u003ePublic-Private Partnership (PPP)\u003c/em\u003e. This new wave of reform aimed not only to expand capacity but also to consolidate services within modern, high-tech hospital complexes, commonly referred to as city hospitals.\u003c/p\u003e\u003cp\u003eProject planning and feasibility assessments began in 2009.\u003csup\u003e1\u003c/sup\u003e Under this model, private consortia would finance, design, and maintain the facilities while the Ministry of Health retained control over medical service delivery. In exchange, the government committed to making availability payments to the firms over 25-year periods, indexed to performance metrics and often accompanied by service volume guarantees. Importantly, many contracts were denominated in foreign currency, introducing long-term fiscal risks for the public sector. In addition, the PPP hospital model adopted performance-based payments for providers, which had been introduced in 2004 as part of HTP [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Physicians working in city hospitals receive a fixed base salary funded from the general budget, along with performance-based bonuses which are calculated based on the volume and complexity of services delivered. Performance payments are often drawn from hospital-level revolving funds (d\u0026ouml;ner sermaye).\u003c/p\u003e\u003cp\u003eThe first city hospitals opened in 2017 in Mersin, Yozgat, and Isparta, selected in part due to historical shortages in specialist care and inpatient capacity. The Turkish government continued to expand this public-private partnership (PPP) model and by 2022, over 20 city hospitals had opened, with additional projects either under construction or in the planning phase. According to the Ministry\u0026rsquo;s listing, four city hospitals opened in 2017, four more in 2018, two in 2019, five in 2020, one in 2021, and a further four by the end of 2022 ([\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e];[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]). Gaziantep City Hospital, a 1,875-bed facility, was completed in late 2023 and officially opened in February 2024, becoming one of the largest health complexes in southeastern T\u0026uuml;rkiye. Additional PPP projects remain under construction or tender evaluation as of 2025. Large metropolitan facilities such as Ankara Bilkent and Istanbul Başakşehir received significant investment and high-capacity healthcare infrastructure.\u003c/p\u003e\u003cp\u003eThe city hospital initiative sought to address allocative inefficiencies by expanding infrastructure in underserved areas and productive inefficiencies by centralizing services to achieve economies of scale. However, concerns persist regarding its capacity to meaningfully improve physical infrastructure and service quality. The guaranteed payments, even when demand is lower than expected, limit fiscal flexibility. Additionally, the private sector bears minimal demand risk, raising questions about incentive alignment and cost-effectiveness. Critics have also noted the potential for PPP hospitals to crowd out investments in primary care or contribute to over-medicalization ([\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e] ;[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e];[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]). Although performance-based physician incentives included in PPP contracts can enhance productivity and reduce patient waiting times, they can also promote overuse, especially in systems with limited gatekeeping and relatively low patient co-payments ([\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] ;[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] ). The sharp rise in out-of-pocket health expenditures, from 8.2\u0026nbsp;billion TL in 2009 to 220.9\u0026nbsp;billion TL in 2023 ([\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e];[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]), contributes to concerns about overutilization and patient welfare. Assessing the impact of these hospitals on healthcare utilization and health outcomes is therefore essential for understanding the trade-offs inherent in such large-scale infrastructure reforms.\u003c/p\u003e"},{"header":"3. Literature Review","content":"\u003cp\u003eOver the past decade, governments of both developing and developed countries have increasingly implemented public-private partnerships (PPPs) in health care, with the goal of improving the delivery of health care in a cost-effective manner and ultimately improving health outcomes. Despite this growing policy relevance, the literature on PPPs in health care is comparatively underdeveloped [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eSeveral studies rely on descriptive comparisons\u0026mdash;such as contrasting PPP and public hospitals or assessing outcomes before and after PPP adoption\u0026mdash;to evaluate their effectiveness. The findings are mixed. Evidence from Spain\u0026rsquo;s Alzira model suggests no consistent advantage over public hospitals in technical efficiency or unit costs [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] and [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. In Lesotho, a PPP hospital network outperformed government-run facilities on clinical measures but incurred significant cost overruns [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. In Uganda, a PPP hospital expanded outreach and facility-based deliveries but left some services unaffordable [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Evaluations of India\u0026rsquo;s Rashtriya Swasthya Bima Yojana (RSBY) and its successor Ayushman Bharat (PM-JAY) highlight contractual breaches and \u0026ldquo;cream skimming,\u0026rdquo; with private providers favoring low-cost, high-margin services, problems attributed to weak regulation and monitoring [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. Overall, the literature shows no clear consensus on the impact of PPPs in health care, reflecting wide variation in how they are designed, implemented, and assessed [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Moreover, most descriptive studies cannot rule out alternative explanations, such as secular trends, for observed effects.\u003c/p\u003e\u003cp\u003eFabre and Straub [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e] focus on studies evaluating PPPs in the health sector that employ causal inference methods such as difference-in-differences. Overall, the evidence is mixed, much like the findings from descriptive studies. For example, Mohanan et al. [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e] found that India\u0026rsquo;s Chiranjeevi Yojana program had no impact on institutional delivery rates or maternal outcomes, possibly because the quality of participating facilities was perceived as low or because private providers were still charging fees for facility delivery. In Afghanistan, Engineer et al. [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] evaluated a pay-for-performance program designed to improve maternal and child health services and found no overall effect, partly because health workers did not fully understand the payment incentives. However, they did observe increases in patient interaction time and quality of care. In Congo, Huillery and Seban [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e] found that a PPP incentivizing providers to increase utilization paradoxically reduced both utilization and provider revenue, driven not by lower effort but by patients\u0026rsquo; perception of reduced care quality. By contrast, Soeters et al. [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] evaluated another performance-based program in Congo, which rewarded both service utilization and care quality, and found improvements in financial access, facility revenues, and quality of care. Cristia et al. [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] report that a PPP involving mobile medical teams in Guatemala substantially increased children\u0026rsquo;s immunization rates and shifted prenatal care provision from traditional midwives to physicians and nurses. In Cameroon, De Walque et al. [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] found that a facility-based program raised utilization for some services, such as vaccinations and family planning, but not for others such as antenatal care and facility-based deliveries. In Kenya, Obare et al. [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e] showed that a reproductive health voucher program increased institutional deliveries and skilled birth attendance, though it had no effect on antenatal care. Synthesizing these findings, Fabre and Straub [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e] conclude that PPPs are more effective when incentive payments are tied to both service volume and quality, and when demand- and supply-side policies are combined to improve utilization and outcomes.\u003c/p\u003e\u003cp\u003eGiven the mixed findings, further research is needed to better understand the conditions under which PPPs improve access and health outcomes. Our study contributes to this literature by providing the first evaluation of T\u0026uuml;rkiye\u0026rsquo;s city hospital reform, one of the largest PPP initiatives undertaken in the health sector. We employ a difference-in-differences approach to identify causal effects, thereby contributing new evidence on the impact of this reform and more generally, on the impact of PPPs in the health care sector.\u003c/p\u003e"},{"header":"4. Data and Empirical Methodology","content":"\u003cp\u003eWe use repeated cross-sectional data from the 2008, 2010, 2014, 2016, 2019, and 2022 waves of the T\u0026uuml;rkiye Health Survey, a nationally representative survey administered by the Turkish Statistical Institute. The THS collects comprehensive data on individuals' healthcare use, chronic illnesses, perceived health, health practices, and sociodemographic background. The sample design guarantees comparability across survey waves and representativeness at the NUTS-2 regional level.\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eOur dependent variables include binary indicators for any inpatient or outpatient service use in the past year, and an indicator for self-reported bad or very bad health (very good, good, and fair health form the reference category). Independent variables include binary indicators for female, age categories (25\u0026ndash;44 years, 45\u0026ndash;54 years, 55\u0026ndash;64 years, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\ge\\:\\)\u003c/span\u003e\u003c/span\u003e 65 years, with \u0026lt;\u0026thinsp;25 years as the reference category), married status (the reference category includes widowed, divorced, and never married), income categories (middle and high with poor as the reference group), missing income, and education level (secondary and tertiary with primary as the reference group). Household income was reported in categorical brackets that varied across survey years (5 categories in 2014\u0026ndash;2016, 20 categories in 2019\u0026ndash;2022). To ensure comparability across survey waves, we collapsed income into three groups that were consistently defined across waves: low income (0\u0026ndash;1,200 TL or approximately 450\u0026ndash;500 USD), middle income (1,200\u0026ndash;2,500 TL or approximately 500\u0026ndash;850 USD), and high income (\u0026ge;\u0026thinsp;2,500 TL or approximately 850 USD+). Missing responses were coded as a separate dummy variable.\u003c/p\u003e\u003cp\u003eWe categorize individuals as belonging to the treatment or control group based on their NUTS-2 region of residence, which is the most granular geographic identifier available in the survey. The control group includes individuals living in areas without any city hospital openings throughout our study period (2008\u0026ndash;2022), and the treatment group includes individuals living in NUTS-2 areas where city hospitals were opened in 2017 or 2018. We exclude NUTS-2 regions where city hospitals were opened after 2018 to allow a sufficient post period necessary for the difference-in-differences analysis (described below) and to avoid bias in two-way fixed effects models due to the staggered rollout and heterogeneity in treatment effects [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the distribution of treatment, control, and excluded NUTS-2 regions across T\u0026uuml;rkiye. Treatment regions are depicted in red, control regions in light pink, and excluded regions are in gray.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAfter excluding observations with missing or inconsistent values and in regions where city hospitals opened after 2018, our analysis sample includes 115,673 observations, with 27,460 in treatment regions and the remaining in control regions. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents summary statistics for health care use, health outcomes, and sociodemographic characteristics for our analysis sample, and separately for the treatment and control groups. Nearly 54% of the sample is female, 13% are aged 65 or older, and 13.7% have completed tertiary education. About 10% report any inpatient use in the past 12 months, while 63% report using outpatient care. The sample is relatively healthy, with fewer than 10% reporting bad or very bad health. The treatment and control groups are similar across most observable characteristics, although the control group has slightly higher income and educational attainment. Importantly, our empirical strategy is a difference-in-differences design, which does not require balance between the two groups, but rather relies on the assumption of parallel trends.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eSummary Statistics\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(1)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(3)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFull Sample\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eControl Group\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eTreatment Group\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAny Inpatient Use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.101\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.101\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.102\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.302)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.301)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.303)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAny Outpatient Use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.633\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.632\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.636\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.482)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.482)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.481)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBad or Very Bad Health\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.096\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.089\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.118\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.294)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.284)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.323)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.538\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.538\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.538\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.499)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.499)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.499)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMarried\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.685\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.684\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.691\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.464)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.465)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.462)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge 25\u0026ndash;44\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.391\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.393\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.385\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.488)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.488)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.487)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge 45\u0026ndash;54\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.175\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.176\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.173\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.380)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.380)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.378)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge 55\u0026ndash;64\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.133\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.132\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.137\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.340)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.339)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.344)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.130\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.130\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.131\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.337)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.337)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.337)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMiddle Income\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.387\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.388\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.384\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.487)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.487)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.486)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHigh Income\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.220\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.230\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.186\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.414)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.421)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.389)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMissing Income\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.259\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.259\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.260\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.438)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.438)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.439)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSecondary Education\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.142\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.144\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.135\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.349)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.351)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.342)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTertiary Education\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.137\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.140\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.125\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.344)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.347)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.331)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eObservations\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e115,673\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e88,213\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e27,460\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"4\"\u003eSource: T\u0026uuml;rkiye Health Survey Data (2008\u0026ndash;2022). Notes: Table presents means and standard deviations are in parentheses.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eTo identify the causal effect of the reform on health care use and health outcomes, we employ the following difference-in-differences regression model:\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Y}_{it}={\\beta\\:}_{0}+{\\beta\\:}_{1}Treatmen{t}_{i}\\times\\:Pos{t}_{t}+{\\beta\\:}_{2}Treatmen{t}_{i}+{\\beta\\:}_{3}Pos{t}_{t}+{\\beta\\:}_{4}{X}_{it}+{\\epsilon\\:}_{it}\\)\u003c/span\u003e\u003c/span\u003e (Eq.\u0026nbsp;1)\u003c/p\u003e\u003cp\u003eWhere, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Y}_{it}\\)\u003c/span\u003e\u003c/span\u003e represents inpatient care use, outpatient care use, or health outcomes for person \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\:\\)\u003c/span\u003e\u003c/span\u003ein survey year \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:t\\)\u003c/span\u003e\u003c/span\u003e. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Treatmen{t}_{i}\\)\u003c/span\u003e\u003c/span\u003e is one if person \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e resides in a region with a city hospital opening and zero if they reside in a region without any city hospital openings during the study period. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Pos{t}_{t}\\)\u003c/span\u003e\u003c/span\u003e is one for observations in 2019 and 2022 and zero for observations in 2008, 2010, 2014, or 2016. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{X}_{it}\\)\u003c/span\u003e\u003c/span\u003e is a vector of sociodemographic characteristics (gender, age, marital status, income, and education) described above. In addition, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{X}_{it}\\)\u003c/span\u003e\u003c/span\u003e includes region-specific linear trends which account for unobserved, time-varying factors that may be correlated with differential trends in health care use and health outcomes. Following the discussion in Abadie et al. [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e], standard errors are clustered at the treatment assignment level which is the NUTS-2 region level in our application. Since there are only 20 NUTS-2 regions in our sample, we use a wild cluster bootstrap with Webb weights to adjust for possible downward bias in standard errors originating from a small number of clusters [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe coefficient of interest is \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\beta\\:}_{1}\\)\u003c/span\u003e\u003c/span\u003e, which identifies the causal effect of the reform on outcomes under the parallel trends assumption. With the inclusion of region-specific linear trends, identification relies on the assumption that, in the absence of the reform, any differences between treated and control groups follow parallel \u003cem\u003enonlinear\u003c/em\u003e trends. This is less restrictive than the standard parallel trends assumption, as it allows for differential linear trends across regions.\u003c/p\u003e\u003cp\u003eTo assess the validity of this assumption, we use an event study regression model which replaces the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Pos{t}_{t}\\)\u003c/span\u003e\u003c/span\u003e variable in Eq.\u0026nbsp;(1) with a full set of year fixed effects.\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Y}_{it}={\\beta\\:}_{0}+\\sum\\:_{t}{\\beta\\:}_{1t}Treatmen{t}_{i}\\times\\:Yea{r}_{t}+{\\mu\\:}_{g}+{\\sum\\:_{t}{\\beta\\:}_{2t}Year}_{t}+{\\beta\\:}_{3}{X}_{it}+{\\epsilon\\:}_{it}\\)\u003c/span\u003e\u003c/span\u003e (Eq.\u0026nbsp;2)\u003c/p\u003e\u003cp\u003eWhere, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Yea{r}_{t}\\)\u003c/span\u003e\u003c/span\u003e represents year fixed effects (with 2016 as the reference year), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\mu\\:}_{g}\\)\u003c/span\u003e\u003c/span\u003e represents region fixed effects,\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\:\\)\u003c/span\u003e\u003c/span\u003eand all other terms are defined as above, except that \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{X}_{it}\\)\u003c/span\u003e\u003c/span\u003e does not include region specific linear trends. A lack of differential trends between treatment and control regions in the pre period (2008\u0026ndash;2016) would provide support for the identifying assumption. As a further check of our identification strategy, we conduct a placebo test by splitting the control group into two arbitrary subsets, designating one as a \u0026ldquo;placebo treatment\u0026rdquo; group and the other as a \u0026ldquo;placebo control\u0026rdquo; group and re-estimating the difference-in-differences model using only control observations. Since none of the observations in the control group were exposed to the opening of a city hospital during the study period, the difference-in-differences estimate should be statistically indistinguishable from zero if the identifying assumption is satisfied. On the other hand, a significant effect would be an indication of unobserved differential trends across regions which may explain the main estimates.\u003c/p\u003e"},{"header":"5. Results","content":"\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents the difference-in-differences estimates of the impact of city hospital openings on inpatient and outpatient healthcare use and on self-reported health. Odd numbered columns present results from a specification that does not include any covariates while even numbered columns present results from a specification that includes sociodemographic characteristics and region-specific linear trends. Wild cluster bootstrap p-values are presented in brackets and are used for inference. For reference, we also present conventional robust standard errors clustered at the NUTS-2 region level. Including covariates improves the precision of the difference-in-differences estimates, and therefore, we focus on the specification with covariates as our preferred specification. We find that city hospital openings significantly increased inpatient care use and improved self-reported health. Specifically, there is 3.7 percentage point increase in the probability of using any inpatient care in response to the reform (Column 2). This represents a 36.6% increase relative to the sample mean. The use of outpatient care also increased by 7.3 percentage points, but this estimate is only significant at the 10% level (Column 4). The program reduced reports of bad or very bad health by 3.9 percentage points (Column 6), which represents a 40.6% relative decrease.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eEffect of City Hospitals on Health Care Use and Health Outcomes\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(1)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(3)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(4)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(5)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003e(6)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAny Inpatient Use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAny Inpatient Use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAny Outpatient Use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eAny Outpatient Use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eBad or Very Bad Health\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eBad or Very Bad Health\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTreatment X Post\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.014**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.037***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.011\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.073*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-0.017*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-0.039***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.005)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.006)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.028)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(0.041)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(0.009)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e(0.010)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e[0.009]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e[0.000]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e[0.732]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e[0.287]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e[0.169]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e[0.010]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePost\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.023***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.041***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.074***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-0.010**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.013\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.003)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.007)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.010)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(0.015)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(0.004)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e(0.009)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e[0.957]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e[0.003]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e[0.022]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e[0.001]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e[0.060]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e[0.175]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTreatment\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.003\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.014**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.060***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.034***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-0.000\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.006)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.006)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.021)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(0.017)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(0.012)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e(0.011)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e[0.734]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e[0.081]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e[0.977]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e[0.003]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e[0.021]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e[0.987]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCovariates\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eX\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eX\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eX\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e115,618\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e115,618\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e115,670\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e115,670\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e115,651\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e115,651\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAdjusted R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.022\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.002\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.046\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.002\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.121\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"7\"\u003eSource: T\u0026uuml;rkiye Health Survey Data (2008\u0026ndash;2022). Notes: Covariates include gender, age, marital status, income, and educational attainment. Robust standard errors clustered at the NUTS2 level are in parentheses. Wild bootstrap p-values based on 9,999 replications are in brackets. *p\u0026thinsp;\u0026lt;\u0026thinsp;0.10 **p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 ***p\u0026thinsp;\u0026lt;\u0026thinsp;0.001\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eNext, we examine the validity of the identifying assumption using the event study regression. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents the coefficients on the interaction between the treatment dummy and year fixed effects, together with 95% confidence intervals. The interaction for 2016 serves as the reference category. For inpatient use, the pre-trend is relatively flat, with a clear increase in utilization in the 2019 and 2022 waves relative to 2016, consistent with a treatment effect emerging only after the reform. For outpatient use, the pre-trend is similarly flat; there is little change in 2019, followed by a modest uptick that emerges in 2022. For self-reported health, the pre-trend shows an upward trend that is not statistically significant, followed by a downward trend in the post-year period. Overall, the results provide strong support for the assumption of parallel trends for inpatient and outpatient use. In the case of self-reported health, while the graph suggests that the difference between treatment and control regions is increasing over the pre-period, the lack of statistical significance implies that we cannot rule out parallel trends. Moreover, the inclusion of region-specific linear trends in our preferred specification suggests that the identifying assumption of parallel non-linear trends is likely to be satisfied.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAs a further check on the identifying assumptions, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents results from the placebo regressions. For this analysis, the sample is restricted to individuals residing in control regions where no city hospital was constructed during our study period. Since these individuals do not have access to a city hospital, we do not expect the reform to affect their health care use or health outcomes. Thus, any statistically significant effects would indicate spurious treatment effects and a potential violation of the identifying assumptions. Reassuringly, none of the placebo difference-in-differences estimates for inpatient use, outpatient use, or self-reported health are statistically significant. Moreover, the placebo estimates have the opposite sign of our main difference-in-differences estimates, further suggesting that the main results are not driven by unobserved factors contributing to differential trends between treatment and control regions. Together with the evidence from the event study regressions, these findings provide strong support for the validity of the identifying assumptions.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003ePlacebo Tests\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(1)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(3)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAny Inpatient Use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAny Outpatient Use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eBad or Very Bad Health\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePlacebo Treatment X Post\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.003\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.028\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.011)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.041)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.020)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e[0.786]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e[0.988]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e[0.242]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePost\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.040***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.045**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.041***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.007)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.018)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.007)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e[0.005]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e[0.008]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e[0.005]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePlacebo Treatment\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.022*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.049*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.035*\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(0.010)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.027)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.020)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e[0.080]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e[0.108]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e[0.129]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e88,178\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e88,210\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e88,198\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAdjusted R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.022\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.046\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.110\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"4\"\u003eSource: T\u0026uuml;rkiye Health Survey Data (2008\u0026ndash;2022). Notes: All regressions include gender, age, marital status, income, and educational attainment as covariates. Robust standard errors clustered at the NUTS2 level are in parentheses. Wild bootstrap p-values based on 9,999 replications are in brackets. *p\u0026thinsp;\u0026lt;\u0026thinsp;0.10 **p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 ***p\u0026thinsp;\u0026lt;\u0026thinsp;0.001\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"6. Conclusion and Discussion","content":"\u003cp\u003eUsing nationally representative data and a difference-in-differences methodology, we find that T\u0026uuml;rkiye\u0026rsquo;s city hospital reform significantly increased inpatient utilization by 3.7 percentage points, which represents a sizeable relative increase of 36.6% given the low rates of inpatient utilization in this population. There is some evidence of an increase in outpatient utilization, but the effect is only significant at the 10% level. We also find a 3.9 percentage point decrease in self-reports of bad or very bad health, which represents a 40.6% relative decline. The increase in utilization combined with health improvements suggests that the reform meaningfully expanded access to hospital services and mitigates concerns about physician induced demand or the overuse of care. Our findings are broadly consistent with studies showing that PPPs can improve service delivery when performance-based incentives are linked to utilization and quality outcomes [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e], [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. However, the results differ from evidence in countries such as India and Afghanistan, where incentive programs had limited effects due to weak implementation or low perceived quality [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. The Turkish experience shows that combining infrastructure expansion with performance-linked physician incentives can meaningfully enhance access and health outcomes, particularly when capacity constraints are more severe.\u003c/p\u003e\u003cp\u003eOur study offers important insights into how infrastructure policy and provider incentives interact in shaping health capital. The findings suggest that building tertiary care facilities, particularly in underdeveloped areas, can enhance service utilization when paired with performance-based incentives to increase provider availability and effort. However, due to data constraints, our analysis is limited to a select set of utilization and health measures and covers only two years of the post-reform period. To fully assess the impact of this reform, future research should examine longer-term changes in access and health, incorporate more detailed measures of utilization and health outcomes, and explore the effects of expanding city hospitals into other regions, which may face different challenges or experience varying impacts.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003cp\u003eThis study uses de-identified secondary microdata from the Turkish Statistical Institute (TURKSTAT, T\u0026Uuml;İK) T\u0026uuml;rkiye Health Survey (THS) 2008\u0026ndash;2022, accessed under a formal data use permission. The data contain no personally identifiable information, and the analyses involved no direct contact with human participants or experimental interventions. All procedures were performed in accordance with the Turkish Statistical Law No. 5429 and TURKSTAT\u0026rsquo;s confidentiality and data security standards. Therefore, the study is exempt from institutional ethics review, and no additional ethical approval was required. Informed consent was obtained by TURKSTAT during the data collection process.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003cp\u003eNot applicable. The study uses secondary anonymized data that do not contain identifiable individual information.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003ch2\u003eCompeting interests\u003c/h2\u003e\u003cp\u003eThe authors declare that they have no financial, non-financial, business, or family competing interests.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eThis research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. However, the first author acknowledges support from the Scientific and Technological Research Council of T\u0026uuml;rkiye (T\u0026Uuml;BİTAK) 2219 International Postdoctoral Research Fellowship Program.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eDTK (Kirklareli University, University of South Florida): Conceptualization, data curation, formal analysis, methodology, writing original draft preparation. PA (University of South Florida): Supervision, validation, writing, review \u0026amp; editing. All authors read and approved the final manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe author thanks the Scientific and Technological Research Council of T\u0026uuml;rkiye (T\u0026Uuml;BİTAK).\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe data supporting the findings of this study are taken from the T\u0026uuml;rkiye Health Survey (THS) conducted by the Turkish Statistical Institute (TURKSTAT). Due to data confidentiality policies, these microdata cannot be shared publicly by the authors. Researchers may request access through the official TURKSTAT microdata application process:[https://www.tuik.gov.tr/Kurumsal/Mikro\\_Veri](https:/www.tuik.gov.tr/Kurumsal/Mikro_Veri) .\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eFabre A, Straub S. The Impact of Public\u0026ndash;Private Partnerships (PPPs) in Infrastructure, Health, and Education. J Econ Lit. 2023;61:655\u0026ndash;715. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1257/JEL.20211607\u003c/span\u003e\u003cspan address=\"10.1257/JEL.20211607\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcguire TG. Handbook of Health Economics. In: Culyer AJ, Newhouse JP, editors. Handbook of Health Economics. 1A edition. Amsterdam: Elsevier Science B.V.; 2000. pp. 461\u0026ndash;536. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S1574-0064(00)80168-7\u003c/span\u003e\u003cspan address=\"10.1016/S1574-0064(00)80168-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePropper C, Reenen J, Van. Can Pay Regulation Kill? Panel Data Evidence on the Effect of Labor Markets on Hospital Performance. J Polit Econ. 2010;118:189\u0026ndash;408. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1086/653137\u003c/span\u003e\u003cspan address=\"10.1086/653137\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTatar M, Mollahaliloǧlu S, Sahin B, Aydin S, Maresso A, Hern\u0026aacute;ndez-Quevedo C. Turkey Health system review. Health Syst Transit. 2011;13:1\u0026ndash;186, xiii.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTurkish Ministry of Health, (Ed. Prof. Dr. Recep Akdağ). Turkey Health Transformation Program (2003\u0026ndash;2010): Evaluation Report. Ankara; 2011.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOECD. Health at a Glance 2023: OECD Indicators. Paris: OECD Publishing; 2023. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1787/7a7afb35-en\u003c/span\u003e\u003cspan address=\"10.1787/7a7afb35-en\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAktas P, Hammond J, Richardson L. Pay-for-performance in healthcare provision: the role of discretion in policy implementation in Turkey. Int J Public Sect Manag. 2023;36:530\u0026ndash;45. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1108/IJPSM-12-2022-0282\u003c/span\u003e\u003cspan address=\"10.1108/IJPSM-12-2022-0282\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMinistry of Health of Turkey Directorate General of Health Investments. Ten city hospitals opened in three years. 2020. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://sygm.saglik.gov.tr/TR,58066/3-yilda-10-sehir-hastanesi-hizmete-acildi.html\u003c/span\u003e\u003cspan address=\"https://sygm.saglik.gov.tr/TR,58066/3-yilda-10-sehir-hastanesi-hizmete-acildi.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 2 Aug 2025.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAli D, Benli RSD, Platform. The modern face of Turkey: City hospitals. 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://sdplatform.com/turkiyenin-modern-yuzu-sehir-hastaneleri/#:~:text=Adana%2C\u003c/span\u003e\u003cspan address=\"https://sdplatform.com/turkiyenin-modern-yuzu-sehir-hastaneleri/#:~:text=Adana%2C\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Elazığ%2C Eskişehir%2C Isparta,adet şehir hastanesi işletmeye a\u0026ccedil;ılmıştır. Accessed 30 Jul 2025.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAri HO, Şahi̇n B. Evaluation of health personnel\u0026rsquo;s views on city hospitals (CH\u0026rsquo;s) as a public-private partnership practice in the Turkish health system. Int J Healthc Manag. 2024;17:756\u0026ndash;65. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/20479700.2023.2235790\u003c/span\u003e\u003cspan address=\"10.1080/20479700.2023.2235790\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDemirag I, Fırtın CE, Tekin Bilbil E. Managing expectations with emotional accountability: making City Hospitals accountable during the COVID-19 pandemic in Turkey. J Public Budg Acc Financ Manag. 2020;32:889\u0026ndash;901. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1108/JPBAFM-07-2020-0097\u003c/span\u003e\u003cspan address=\"10.1108/JPBAFM-07-2020-0097\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e\u0026Ouml;zzeybek Taş M, Tengilimoglu D, Atilla EA, A Research For Determining The Opinions On City Hospitals Example Of Ankara. The Capital City. A Res Determ Opin city Hosp Ex Ankara. Cap city. 2019;4:140\u0026ndash;59. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.31201/IJHMT.614614\u003c/span\u003e\u003cspan address=\"10.31201/IJHMT.614614\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLee DW, Jang J, Choi DW, Jang SI, Park EC. The effect of shifting medical coverage from National Health Insurance to Medical Aid type I and type II on health care utilization and out-of-pocket spending in South Korea. BMC Health Serv Res. 2020;20:1\u0026ndash;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12913-020-05778-2\u003c/span\u003e\u003cspan address=\"10.1186/s12913-020-05778-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTurkish Statistical Institute (TURKSTAT). 2023 Health Expenditure Statistics of T\u0026uuml;rkiye. Turkish Statistical Institute News. 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://data.tuik.gov.tr/Bulten/Index?p=Saglik-Harcamalari-Istatistikleri-2023-53561\u003c/span\u003e\u003cspan address=\"https://data.tuik.gov.tr/Bulten/Index?p=Saglik-Harcamalari-Istatistikleri-2023-53561\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 2 Aug 2025.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTurkish Statistical Institute (TURKSTAT). 2009\u0026ndash;2012 Health Expenditure Statistics of T\u0026uuml;rkiye, Turkish Statistical Institute News. 2013. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://data.tuik.gov.tr/Bulten/Index?p=Health-Expenditure-Statistics-2009-2012-15871\u003c/span\u003e\u003cspan address=\"https://data.tuik.gov.tr/Bulten/Index?p=Health-Expenditure-Statistics-2009-2012-15871\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 2 Aug 2025.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eComendeiro-Maal\u0026oslash;e M, Ridao-L\u0026oacute;pez M, Gorgemans S, Bernal-Delgado E. Public-private partnerships in the Spanish National Health System: The reversion of the Alzira model. Health Policy (New York). 2019;123:408\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.healthpol.2019.01.012\u003c/span\u003e\u003cspan address=\"10.1016/j.healthpol.2019.01.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCaballer-Tarazona M, Vivas-Consuelo D. A cost and performance comparison of Public Private Partnership and public hospitals in Spain. Health Econ Rev. 2016;6:1\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13561-016-0095-5\u003c/span\u003e\u003cspan address=\"10.1186/s13561-016-0095-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcIntosh N, Grabowski A, Jack B, Nkabane-Nkholongo EL, Vian T. A public-private partnership improves clinical performance in a hospital network in Lesotho. Health Aff. 2015;34:954\u0026ndash;62. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1377/hlthaff.2014.0945\u003c/span\u003e\u003cspan address=\"10.1377/hlthaff.2014.0945\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAsasira J, Ahimbisibwe F. Public-Private Partnership in Health Care and Its Impact on Health Outcomes: Evidence from Ruharo Mission Hospital in Uganda. Int J Soc Sci Stud. 2018;6:79. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.11114/ijsss.v6i12.3911\u003c/span\u003e\u003cspan address=\"10.11114/ijsss.v6i12.3911\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKhetrapal S, Acharya A, Mills A. Assessment of the public-private-partnerships model of a national health insurance scheme in India. Soc Sci Med. 2019;243. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.socscimed.2019.112634\u003c/span\u003e\u003cspan address=\"10.1016/j.socscimed.2019.112634\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRodrigues NJP. Public\u0026ndash;Private Partnerships Model Applied to Hospitals\u0026mdash;A. Crit Rev Healthc. 2023;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/healthcare11121723\u003c/span\u003e\u003cspan address=\"10.3390/healthcare11121723\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMohanan M, Bauhoff S, La Forgia G, Babiarz KS, Singh K, Miller G. Effect of Chiranjeevi Yojana on institutional deliveries and neonatal and maternal outcomes in Gujarat, India: a difference-in-differences analysis. Bull World Heal Organ. 2014;92:187\u0026ndash;94. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2471/BLT.13.124644\u003c/span\u003e\u003cspan address=\"10.2471/BLT.13.124644\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEngineer CY, Dale E, Agarwal A, Agarwal A, Alonge O, Edward A, et al. Effectiveness of a pay-for-performance intervention to improve maternal and child health services in Afghanistan: a cluster-randomized trial. Int J Epidemiol. 2016;45:451\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/IJE/DYV362\u003c/span\u003e\u003cspan address=\"10.1093/IJE/DYV362\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuillery E, Seban J. Financial incentives, efforts, and performances in the health sector: Experimental evidence from the Democratic Republic of Congo. Econ Dev Cult Change. 2021;69. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1086/703235\u003c/span\u003e\u003cspan address=\"10.1086/703235\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSoeters R, Peerenboom PB, Mushagalusa P, Kimanuka C. Performance-Based Financing Experiment Improved Health Care In The Democratic Republic Of Congo. Health Aff. 2011;30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1377/hlthaff.2009.0019\u003c/span\u003e\u003cspan address=\"10.1377/hlthaff.2009.0019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCristia J, Evans WN, Kim B. Improving the Health Coverage of the Rural Poor: Does Contracting-Out Mobile Medical Teams Work? J Dev Stud. 2015;51:247\u0026ndash;61. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00220388.2014.976617\u003c/span\u003e\u003cspan address=\"10.1080/00220388.2014.976617\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDe Walque D, Robyn PJ, Saidou H, Sorgho G, Steenland M. Looking into the performance-based financing black box: evidence from an impact evaluation in the health sector in Cameroon. Health Policy Plan. 2021;36:835. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/heapol/czab002\u003c/span\u003e\u003cspan address=\"10.1093/heapol/czab002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eObare F, Warren C, Njuki R, Abuya T, Sunday J, Askew I, et al. Community-level impact of the reproductive health vouchers programme on service utilization in Kenya. Health Policy Plan. 2013;28:165\u0026ndash;75. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/HEAPOL/CZS033\u003c/span\u003e\u003cspan address=\"10.1093/HEAPOL/CZS033\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGoodman-Bacon A. Difference-in-differences with variation in treatment timing. J Econom. 2021;225:254\u0026ndash;77. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.JECONOM.2021.03.014\u003c/span\u003e\u003cspan address=\"10.1016/J.JECONOM.2021.03.014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbadie A, Athey S, Imbens GW, Wooldridge JM. When Should You Adjust Standard Errors for Clustering? Q J Econ. 2022;138:1\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/QJE/QJAC038\u003c/span\u003e\u003cspan address=\"10.1093/QJE/QJAC038\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eColin Cameron A, Miller DL. A Practitioner\u0026rsquo;s Guide to Cluster-Robust Inference. J Hum Resour. 2015;50:317\u0026ndash;72. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3368/JHR.50.2.317\u003c/span\u003e\u003cspan address=\"10.3368/JHR.50.2.317\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRepublic of Turkey Official Gazette. Law No. 6428: Law on the construction of facilities through public\u0026ndash;private partnerships in the field of healthcare services and amendments to certain laws and statutory decrees. Ankara; 2013.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Footnotes","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003e These efforts were formalized through the passage of \u003cem\u003eLaw No. 6428\u003c/em\u003e in 2013, enabling PPPs in health infrastructure [32].\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e NUTS-2 regions refer to the second level administrative regions according to Nomenclature of Territorial Units for Statistics. There are 26 NUTS-2 regions in T\u0026uuml;rkiye.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"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":"Inpatient utilization, outpatient utilization, self-reported health, public-private partnerships, hospital infrastructure, difference-in-differences","lastPublishedDoi":"10.21203/rs.3.rs-7970303/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7970303/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study examines the impact of T\u0026uuml;rkiye\u0026rsquo;s city hospital reform on health care use and health outcomes. As part of a broader health care reform, the Turkish government established public-private partnerships to upgrade existing hospital infrastructure and build new high-capacity tertiary facilities across selected regions of the country. The first set of hospitals began operations in 2017. To estimate the effect of this reform, we employ a difference-in-differences approach, comparing changes in outcomes before and after 2017\u0026ndash;18 in regions that opened city hospitals (treatment regions) with regions that did not open any city hospitals during the study period (control regions). We find that the reform led to a significant increase in inpatient utilization and improvements in self-reported health. Specifically, there is a 3.7 percentage point increase in the probability of using any inpatient care and 3.9 percentage point decrease in the probability of reporting bad or very bad health.\u003c/p\u003e","manuscriptTitle":"The Impact of Türkiye’s City Hospitals on Health Care Use and Health Outcomes","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-17 11:19:04","doi":"10.21203/rs.3.rs-7970303/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":"f44198b0-04cd-4b63-8c75-7cb2f6a1540e","owner":[],"postedDate":"November 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-11-17T11:19:07+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-17 11:19:04","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7970303","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7970303","identity":"rs-7970303","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}