The political economy of health reforms and out-of-pocket spending in Türkiye: evidence from household microdata, 2014–2023

Abstract

Background

Out-of-pocket (OOP) health spending remains a major challenge to financial protection and health equity in many middle-income countries, even in contexts where insurance coverage has expanded. Türkiye represents a salient case, as large-scale health reforms over the past two decades have coincided with persistent OOP burdens and increasing macroeconomic volatility.

Methods

This study analyzes nationally representative microdata from the Turkish Household Budget Survey covering the period 2014–2023. Log-linear regression models were used to examine the socio-demographic determinants of individual-level out-of-pocket (OOP) health expenditures. To assess equity and financial protection, both the OOP-to-income ratio and a World Health Organization–referenced 10% threshold were employed as descriptive benchmarks, rather than to estimate full population-level catastrophic health expenditure incidence.

Results

OOP health expenditures increased over the study period and remained above OECD averages. While higher-income groups incurred greater absolute spending, the relative financial burden was disproportionately concentrated among lower-income households. After 2018, the concentration of high OOP-to-income burdens increased, particularly among disadvantaged groups, suggesting a weakening of financial protection during periods of economic turbulence and structural reform.

Conclusions

The findings indicate that health reforms in Türkiye expanded access to services but did not prevent the unequal distribution of financial risk across income groups. From a health equity perspective, modernization strategies that rely on public–private partnerships and digital expansion may unintentionally shift cost pressures onto households. Strengthening financial protection mechanisms and systematically monitoring distributional effects are essential for advancing equity-oriented health reform.

Introduction

Transformations in health systems are not limited to technical adjustments in service delivery; rather, they should be understood as part of a broader political–economic process shaped by the interaction of budgetary policies, inter-institutional relations, socio-economic inequalities, and political priorities [12, 20]. In middle-income countries, these dynamics are often intertwined with economic vulnerabilities, fiscal constraints, and rising demand for health services, which in turn intensify the financial burden borne by households through out-of-pocket (OOP) health expenditures [24, 29].

Over the past two decades, Türkiye has undergone a comprehensive reform process in the health sector. Structural changes initiated under the Health Transformation Program have subsequently diversified through the construction of large-scale city hospitals under public–private partnership (PPP) models, the implementation of national digital health platforms, and regulatory reforms in pharmaceutical policies [3, 5, 30]. While these reforms have expanded service coverage and improved accessibility, persistently high levels of household OOP expenditures during the same period have created a notable area of tension. OECD data indicate that Türkiye’s OOP ratios have remained above the OECD average for many years [15], suggesting that the reforms may have produced unequal effects in terms of financial protection.

The level of out-of-pocket health expenditures is shaped not only by individuals’ health service utilization preferences but also by socio-demographic determinants such as income, education, gender, insurance coverage, and household structure, as well as by the nature of reforms, macroeconomic conditions, and structural transformations in health service supply [18, 31]. The growing influence of structural factors—such as digitalization, PPP investments, and budgetary policies—on OOP expenditures has rendered the transformation of Türkiye’s health system more pronounced [14, 28]. Consequently, assessing OOP trends requires moving beyond individual behavior to account for the political–economic context in which health reforms are implemented [17].

This study examines the long-term trajectory of out-of-pocket health expenditures in Türkiye and explores how socio-demographic factors shaping these expenditures interact with different reform periods. The analysis is based on microdata from the Household Budget Survey conducted by the Turkish Statistical Institute (TURKSTAT) for the period 2014–2023 [22]. Multiple regression models are employed to assess the effects of age, gender, education, income, insurance status, and household characteristics on OOP expenditures. By discussing the findings within the political–economic context of Türkiye’s health reforms, the study integrates OOP analyses at both micro and macro levels and offers an original contribution to the existing literature.

Political economy of health reforms

Health reforms are not merely technical interventions but are deeply shaped by political priorities, fiscal constraints, and power relations among key actors within health systems [9, 19, 20]. In middle-income countries, limited public resources, pressures for rapid modernization, and pre-existing socio-economic inequalities strongly influence how reforms are designed and implemented, as well as who ultimately bears their costs [12, 18].

From a political economy perspective, three questions are central to understanding reform outcomes: under what fiscal and political conditions reforms emerge, which actors shape their direction, and how costs and benefits are distributed across social groups [6, 19]. In Türkiye, health reforms implemented since the early 2000s—particularly the expansion of public–private partnership (PPP)–based hospitals, digital health platforms, and cost-containment measures—have been closely intertwined with broader macroeconomic dynamics and budgetary pressures [3, 17, 28].

While these reforms have expanded service capacity and improved access, they have also altered the cost structure of healthcare delivery. Evidence from multiple health systems suggests that technology-intensive investments and the growing role of private providers may increase unit costs and shift financial risk toward households, especially when regulatory and redistributive mechanisms remain limited [4, 10, 24]. Consequently, out-of-pocket health expenditures cannot be fully explained by individual behavior alone but must be interpreted within the broader political–economic context in which health reforms are implemented.

Political economy of health reforms in Türkiye

Health reforms in Türkiye—particularly those implemented after 2003 under the framework of the Health Transformation Program—have generated a profound transformation in the context of the socio-economic structure, public finance, and relations of political power. The political economy of these reforms has been decisive not only in shaping the organization of health services, but also in redefining state–society relations, the social protection regime, and the distributive structure of the public budget. In the related literature, this process has been examined both through the lens of welfare state transformation [17] and in terms of the evolving composition of the public and private sectors [21].

The neoliberal restructuring process that emerged after the 1980s and profoundly affected the Turkish health system. Throughout the 1990s, high inflation, persistent public deficits, and a lack of institutional coordination led to significant inadequacies in both the financing and accessibility of health services [32]. Reports by the World Bank and the International Monetary Fund during this period highlighted inefficiencies in health expenditures, fiscal sustainability problems within social security institutions, and the fragmented nature of the institutional structure in Türkiye [27].

Policies implemented during this period exhibited clear parallels with similar reform trajectories in Latin America and Eastern Europe. Reform initiatives emphasized the separation of financing and service provision across different institutions, the unification of health insurance schemes under a single umbrella, the introduction of family medicine, and tendencies toward the autonomization of public hospitals [20].

The Health Transformation Program launched in 2003 can be interpreted as both a restructuring of centralized political authority and the construction of a new social policy regime [16]. The establishment of the Social Security Institution (SSI) in 2006 consolidated previously fragmented social security schemes under a single institutional framework, enabling the pooling of funds and creating a relatively more equitable opportunity for access to health services [30]. This development represented an important step toward strengthening redistributive mechanisms within public finance.

The family medicine system sought to address the fragmentation of the former health center model by improving the provision of preventive health services at the individual level [3]. At the same time, the performance-based payment system has been subject to criticism for imposing excessive workloads on health workers [21].

The Public Hospital Unions model introduced in 2011 aimed to manage hospitals according to business-oriented principles; however, it was abolished in 2017 due to the high level of bureaucratic burden and administrative complexity it generated [7]. This experience clearly revealed the political risks associated with adopting neoliberal principles in the design of health system governance.

One of the most visible outcomes of the Health Transformation Program has been the expansion of access to health services. Significant improvements have been observed in key health indicators, including reductions in infant mortality, progress in maternal health outcomes, and a substantial increase in hospital utilization rates ([3]; [13]). These developments support analyses suggesting that the government strengthened its political legitimacy through social policy interventions [17].

The share of health expenditures in Gross Domestic Product (GDP) increased between 2002 and 2015 and subsequently followed a stable path (TURKSTAT, 2023). Rising demand for health services placed growing pressure on the health budget. In response, cost-containment measures—such as reference pricing and global budgeting in pharmaceutical and medical device payments—were introduced [32].

During this period, notable changes also occurred in the role of the private sector within health service delivery. City hospitals constructed through public–private partnerships have become one of the most intensively debated issues in the political economy literature. While critics argue that this model imposes long-term rental obligations on the public sector, reduces budgetary flexibility, and increases financial risks [5], the government has framed PPPs as a strategic instrument for the rapid modernization of health infrastructure.

The political economy of health reforms in Türkiye can be examined from multiple perspectives. Reforms have centralized decision-making processes within the health system, thereby increasing political authority over service provision. Through this transformation program, the ruling political power has strengthened its linkages with voters via social assistance and health services. Türkiye’s reform experience reflects what the literature describes as a “hybrid welfare model,” in which market mechanisms and state-led social spending are intertwined [11, 16].

In summary, the political economy of health reforms in Türkiye constitutes a comprehensive process that extends beyond improvements in health indicators or institutional restructuring alone. It has redefined state–society relations, transformed public finance, and reshaped mechanisms of political legitimacy. Both international waves of reform, particularly those associated with neoliberal health policy transformations, including cost-containment strategies, privatization efforts, and the expansion of market-oriented healthcare models. For this reason, Türkiye’s experience offers an important comparative case for understanding global welfare state transformations and health reform policies in developing countries.

Budgetary policies and fiscal constraints

One of the most decisive factors shaping health reforms in middle-income countries is the level of public budgetary resources allocated to health. Fiscal discipline, borrowing capacity, and long-term investment strategies directly influence the direction of reforms. In Türkiye, the shift toward public–private partnership (PPP) investments since the 2010s has enabled rapid infrastructure modernization in the short term, while generating long-term rental obligations for the public budget. This has constrained fiscal flexibility and, in certain service areas, contributed to a partial transfer of the financial burden to households [28]. Accordingly, the impact of fiscal policies on out-of-pocket (OOP) expenditures aligns with the supply-side cost pressures frequently emphasized in the literature.

Reform opportunities created by crises

The COVID-19 pandemic, as in many countries, led to abrupt policy changes in health financing and service delivery in Türkiye. While crises may reduce out-of-pocket (OOP) expenditures in the short term as the state temporarily assumes a greater financial burden, they can generate cost-increasing effects in the medium term due to postponed treatments, shifts toward private sector services, and heightened demand [18]. Accordingly, periods of crisis function as “windows of opportunity” in which reforms accelerate while existing inequalities become more visible.

Political economy of out-of-pocket expenditures and inequalities

International research indicates that in countries with high levels of out-of-pocket expenditures, inequalities related to income, education, insurance coverage, and geographic disparities are more pronounced [2, 24]. In Türkiye, the persistent fact that out-of-pocket spending has remained above the OECD average for many years highlights both the limited inequality-reducing effects of reforms and the fragility of the health system’s financial protection capacity [15]. For this reason, understanding OOP dynamics is not merely an economic issue, but one that lies at the core of broader debates on social justice, inequality, and the welfare state.

The success and societal impacts of health reforms cannot be explained solely through technical designs or financing models; rather, the emergence, implementation, and outcomes of reforms are largely determined by political priorities, institutional structures, budgetary policies, and power relations among different actors. The political economy literature emphasizes that, particularly in low- and middle-income countries, health reforms are shaped by state capacity, fiscal policies, bureaucratic structures, interest groups, international organizations, and the “windows of opportunity” created during periods of crisis [12, 18, 29].

This historical and political–economic perspective on Türkiye’s health reforms provides a critical backdrop for understanding changes in out-of-pocket health expenditures over time. The political context in which reforms are implemented, along with budget constraints, processes of institutional restructuring, and the public–private balance, has directly shaped the extent to which households participate financially in the health system. Accordingly, OOP dynamics should be evaluated not only through socio-demographic determinants, but also in conjunction with the structural transformations that have emerged across reform periods. In the following section, findings derived from microdata of the Household Budget Survey conducted by the Turkish Statistical Institute (TURKSTAT) are presented within this political–economic framework, offering a detailed examination of the mechanisms through which out-of-pocket health expenditures diverge.

Research methodology

This study employs a quantitative research design to examine how out-of-pocket (OOP) health expenditures in Türkiye evolved over the period 2014–2023 and how their socio-demographic determinants interacted with different phases of health system reform. The analysis is based on nationally representative microdata from the Household Budget Survey (HBS) conducted by the Turkish Statistical Institute (TURKSTAT).

Measurement of financial protection and health equity

To operationalize financial protection and health equity, the study adopts a dual-measurement strategy capturing both relative and threshold-based dimensions of household financial burden.

First, a continuous relative burden indicator was constructed by calculating the ratio of annual out-of-pocket (OOP) health expenditures to annual disposable income (OOP/Income). This measure enables a distribution-sensitive assessment of how health spending translates into financial pressure across income groups and over time, which is central to equity-oriented analyses.

Second, a binary indicator of catastrophic health expenditure was defined using a 10% of annual disposable income, consistent with thresholds commonly applied in global monitoring frameworks developed by the World Health Organization and the World Bank. Individuals were classified as experiencing catastrophic health expenditure if their annual OOP health spending exceeded this threshold. This indicator is used as a descriptive benchmark rather than to estimate population-level catastrophic expenditure incidence using a full capacity-to-pay approach. Accordingly, the study does not make causal claims regarding impoverishment but contextualizes observed expenditure patterns within an equity framework and across reform periods.

Data sources and sample

The Household Budget Survey (HBS) is a nationally representative microdata source capturing detailed information on consumption and expenditure behavior in Türkiye. This study utilizes harmonized HBS microdata from the years 2014, 2017, 2019, and 2023. To provide a clearer overview of the sample structure, descriptive statistics on the distribution of key socio-demographic characteristics are presented in Table 1. The table reports the distribution of observations across gender, marital status, education level, age groups, and insurance status, as well as the total number of observations for each year.

Table 1.

Descriptive statistics

	2014					2017					2019					2023
	low	lower	middle	Upper	High	low	lower	middle	Upper	High	low	lower	middle	Upper	High	low	lower	middle	Upper	High
	Gender
Male	0,301	0,514	0,727	0,810	0,805	0,182	0,198	0,473	0,740	0,802	0,178	0,239	0,491	0,726	0,761	0,180	0,204	0,509	0,700	0,793
Female	0,700	0,490	0,270	0,190	0,195	0,818	0,802	0,527	0,263	0,198	0,822	0,761	0,509	0,274	0,239	0,820	0,796	0,491	0,300	0,207
	Marital Status
Single	0,352	0,240	0,154	0,116	0,074	0,346	0,345	0,256	0,152	0,063	0,334	0,357	0,205	0,141	0,077	0,336	0,357	0,195	0,172	0,092
Married	0,540	0,548	0,745	0,821	0,887	0,637	0,624	0,532	0,759	0,879	0,651	0,603	0,581	0,766	0,862	0,646	0,603	0,592	0,742	0,859
Divorced	0,084	0,187	0,073	0,042	0,021	0,007	0,010	0,039	0,029	0,028	0,009	0,017	0,041	0,038	0,033	0,010	0,019	0,040	0,038	0,033
Others	0,023	0,026	0,028	0,020	0,018	0,011	0,021	0,173	0,060	0,030	0,006	0,023	0,172	0,555	0,028	0,009	0,021	0,172	0,047	0,016
	Health Insurance Status
Public	0,611	0,753	0,896	0,927	0,967	0,800	0,746	0,695	0,922	0,953	0,839	0,785	0,797	0,962	0,971	0,871	0,826	0,824	0,952	0,976
No Insurance	0,086	0,070	0,042	0,042	0,021	0,059	0,053	0,064	0,031	0,024	0,006	0,007	0,007	0,002	0,002	0,000	0,000	0,000	0,000	0,000
Banks’ Fund	0,003	0,008	0,001	0,008	0,004	0,001	0,001	0,000	0,002	0,005	0,152	0,203	0,194	0,033	0,016	0,124	0,172	0,173	0,043	0,015
Private	0,000	0,000	0,000	0,000	0,000	0,001	0,001	0,001	0,001	0,002	0,002	0,002	0,001	0,002	0,007	0,003	0,001	0,001	0,003	0,006
General	0,300	0,170	0,061	0,023	0,007	0,135	0,198	0,240	0,045	0,016	0,002	0,003	0,001	0,001	0,005	0,002	0,001	0,002	0,002	0,003
	Education
Primary	0,555	0,642	0,564	0,441	0,248	0,499	0,479	0,614	0,463	0,285	0,460	0,424	0,610	0,403	0,205	0,385	0,385	0,551	0,305	0,185
Secondary	0,395	0,299	0,364	0,379	0,238	0,438	0,451	0,307	0,413	0,315	0,465	0,478	0,295	0,427	0,275	0,518	0,498	0,341	0,449	0,320
Graduate	0,049	0,056	0,071	0,174	0,452	0,062	0,070	0,077	0,120	0,365	0,071	0,095	0,092	0,166	0,453	0,096	0,114	0,105	0,235	0,432
Post-Graduate	0,000	0,003	0,001	0,007	0,063	0,080	0,110	0,150	0,360	0,030	0,040	0,002	0,002	0,004	0,070	0,001	0,003	0,004	0,012	0,063
	Age
0–30	0,441	0,275	0,224	0,184	0,118	0,436	0,447	0,280	0,201	0,114	0,416	0,456	0,228	0,185	0,112	0,421	0,427	0,197	0,199	0,120
31–45	0,261	0,262	0,301	0,368	0,458	0,237	0,263	0,226	0,339	0,427	0,244	0,253	0,231	0,321	0,451	0,252	0,252	0,209	0,380	0,472
46–60	0,131	0,199	0,248	0,278	0,331	0,199	0,171	0,206	0,227	0,326	0,211	0,174	0,222	0,252	0,300	0,207	0,194	0,215	0,225	0,321
60+	0,167	0,265	0,227	0,166	0,093	0,128	0,119	0,289	0,233	0,134	0,129	0,117	0,319	0,242	0,137	0,121	0,127	0,379	0,195	0,088
Sample	24,269					56,703					41,691					27,614

Microdata for the years 2020–2021 were excluded due to comparability limitations arising from disruptions in survey implementation and inconsistencies in harmonized expenditure variables during the COVID-19 period. The selected years allow for a consistent examination of pre- and post-2018 reform dynamics while ensuring methodological comparability across survey waves.

The HBS data structure consists of three linked components:

• Individual level: age, gender, education, marital status, employment status, and insurance coverage.

• Household level: disposable income, household size, and dependency characteristics.

• Consumption level: detailed health expenditure items.

A unified individual-level analytical dataset was constructed by linking household and consumption records to individuals using TURKSTAT’s survey identifiers. The TURKSTAT Household Budget Survey (HBS) is based on a stratified two-stage sampling design, ensuring national representativeness across urban and rural areas of Türkiye. Households are selected through probabilistic sampling procedures, and survey weights are provided to account for sampling design and non-response adjustments. The survey collects detailed information on household income, consumption, and expenditure through face-to-face interviews conducted at regular intervals. This structure allows for a comprehensive assessment of household-level financial behavior.

In this study, sampling weights provided by TURKSTAT were applied in all analyses to ensure that the results are representative of the national population. This approach strengthens the external validity and generalizability of the findings.

Data harmonization process

Substantial differences across survey years in variable naming, coding schemes, and file structures necessitated a systematic harmonization process. The year 2016 was designated as the reference year. Variable names, expenditure codes, and income definitions in all other survey waves were aligned accordingly. Variables not consistently available across all years were excluded to ensure cross-year comparability. In addition, particular attention was paid to ensuring consistency in variable definitions across survey waves. Potential discrepancies arising from changes in survey design were minimized through harmonization procedures. Despite these efforts, minor measurement inconsistencies across years may remain and are acknowledged as a limitation.

Disaggregation of health expenditures

In this study, the subcomponents of health expenditures were disaggregated according to TURKSTAT’s HBS_KOD5 classification:

• 6110: Pharmaceutical expenditures.

• 6121–6139: Medical products and devices.

• 6211–6212: General practitioner and specialist physician services.

• 6220: Dental services.

• 6231–6239: Laboratory and diagnostic imaging services.

• 6300: Hospital services.

Monthly expenditure values at the consumption-item level were converted into annual expenditures and subsequently assigned to the individual level. The dependent variable in this study is the natural logarithm of annual individual-level out-of-pocket (OOP) health expenditures (ln(OOP)). This transformation is used to reduce skewness and allow coefficient interpretation in relative terms.

Although the TURKSTAT Household Budget Survey provides disaggregated information on different categories of health expenditures (e.g., pharmaceuticals, outpatient care, and hospital services), the dependent variable is defined as total OOP health expenditures at the individual level.

The disaggregated expenditure components are used only in the construction of the total OOP measure and are not modeled separately. This choice ensures a consistent and comparable measure of financial burden across individuals and years, in line with the main objective of the study.

Income groups and weighting

Households were classified into income quintiles based on annual disposable income. TURKSTAT sampling weights were applied at both household and individual levels to ensure national representativeness and to account for survey design effects. Weighted analyses were used to assess distributional differences in OOP expenditures and relative financial burden across income groups.

Econometric specification

Multiple linear regression models were estimated to examine the determinants of out-of-pocket health expenditures.

The baseline specification is given by:

The dependent variable is the natural logarithm of annual individual-level OOP health expenditures. Income is also entered in logarithmic form, allowing coefficient estimates to be interpreted as elasticities and reducing the influence of extreme values.

Models were estimated separately for each year to allow coefficients to vary across structurally different reform periods.

To assess whether major structural breaks altered the income gradient of OOP spending, an extended specification was estimated:

The Post2018 indicator captures the period following major macroeconomic and institutional shifts, while PostPandemic represents the post-COVID adjustment period (2022–2023). The interaction term evaluates whether income-related inequalities in OOP spending intensified after 2018.

Regression analyses were conducted using the statsmodels library in Python. Robust standard errors were applied.

Presentation of findings

Descriptive statistics, expenditure trends, and regression results are presented using tables and figures to illustrate temporal changes, income-group disparities, and reform-related structural shifts. As the study relies exclusively on anonymized and publicly available microdata, ethical committee approval was not required.

Findings

This section presents changes in the level of out-of-pocket health expenditures (OOP), their socio-demographic determinants, and the period-specific effects of health reforms in Türkiye, based on microdata from the Household Budget Survey conducted by the Turkish Statistical Institute (TURKSTAT) for the period 2014–2023. During this period, Türkiye’s out-of-pocket spending remained high and, according to OECD Health Statistics (out-of-pocket expenditure as % of current health expenditure), exceeded the OECD average in most years [15]. (see Fig. 1).

Fig. 1.

The Development of Health Expendiures in Türkiye: An Analsis for the Period 2000–2023. (Source: TURKSTAT; OECD Health Statistics)

Figure 1 provides macro-level context using national accounts/OECD series (2000–2023), while the microdata analysis covers 2014–2023. This increase became particularly pronounced after 2018, coinciding with a period characterized by high inflation, exchange rate volatility, and intensified health reform activity, including the expansion of service capacity in PPP-based city hospitals, as summarized in Table 2.

Table 2.

Development of health reforms in Türkiye (2013–2023)

Year	Health Policies	Description	Potential Impact on Out-of-Pocket Spending
2013	City Hospitals: Construction of large-scale city hospitals began under the public–private partnership model.	Construction of large-scale hospitals through public–private partnerships.	May increase access to advanced services; potential rise in private service demand.
2015	E-Nabız System: The E-Nabız system, which allows citizens to digitally access their health records, was launched.	Nationwide personal health record platform enabling patients to access their health data online.	Could reduce unnecessary visits/tests, lowering costs.
2016	Health Quality Standards: Standards were updated and new criteria were introduced to improve the quality of health services. Imported Drug Restrictions and Domestic Production Incentives: Starting from 2016, incentives were implemented to reduce the use of imported products in the pharmaceutical sector and to increase domestic drug production.	National accreditation and quality evaluation framework for hospitals.	May lead to higher quality but also increased operational costs.
2017	Accreditation Systems: Accreditation processes were initiated to improve the service quality of healthcare institutions.	-	-
2018	Health Tourism: Incentives and promotional activities were intensified to enhance Türkiye’s potential in health tourism.	Policies to encourage domestic production of medicines.	Potential to lower medicine prices in the long term.
2020	Pandemic Management: During the COVID-19 pandemic, contact tracing teams were established, and testing and treatment protocols were developed. Vaccination Programs: The procurement and administration of COVID-19 vaccines were carried out rapidly.	Public financing for COVID-19 tests, treatments, and vaccines.	Short-term reduction in patient burden for pandemic-related care.
2021	Digital Health Services: Telemedicine and remote healthcare services were expanded during the pandemic.	-	-
2022	Quality Assessments: Regular quality evaluations were introduced to measure the service quality of healthcare institutions.	Broader availability of private supplemental insurance options.	May reduce some out-of-pocket costs for insured individuals.

(Source: Ministry of Health reports, official announcements, OECD and WHO assessments)

The 2020–2021 entries provide contextual policy background; HBS microdata for 2020–2021 were not included in the empirical analysis due to data availability/compatibility constraints

The rise in the constant term of log-OOP in the regression models further supports the presence of a structural upward trend in household health expenditure levels.

Regression models were estimated separately for each year to allow the relationships between socio-demographic characteristics and out-of-pocket expenditures to vary across reform periods. This approach facilitates a descriptive assessment of how determinants evolve over time rather than imposing constant coefficients across structurally different periods.

As a robustness check, pooled regression models with year fixed effects were also estimated, yielding substantively similar patterns. These results are available upon request.

Out-of-pocket spending trends

Table 3 presents baseline regression coefficients for selected benchmark years representing different reform phases.

Table 3.

Regression Coefficients for Determinants of Out-of-Pocket Health Expenditures (log-transformed)

Variables	2014	2017	2019	2023
Age	0.0014(0.001)***	0.0019 (0.000)***	0.0051(0.001)***	0.0027(0.001)***
ln(Income)	0.0909 (0.007)***	0.0073(0.002)***	0.0061(0.002)***	0.0113(0.003)***
Education	0.2066 (0.013)***	0.1939(0.010)***	0.2248(0.012)***	0.2359(0.017)***
Gender (Female = 1)	0.1591 (0.019)***	0.0760(0.015)***	0.0778(0.018)***	0.0964(0.027)***
Marital status	-0.0294 (0.017)*	0.0167(0.012)	-0.0130(0.015)	0.0076(0.022)
Pension Fund Insurance	-0.2449 (0.104) **	0.1235(0.153)	-0.2388(0.022)***	-0.4316(0.036)***
Private Insurance	-	0.4039(0.173)**	0.0230(0.160)	0.5910(0.187)***
General Health Insurance	-0.0221 (0.025)	-0.2062(0.017)***	0.4725(0.182)***	0.3258(0.215)
Uninsured	0.0114 (0.038)	0.1337(0.030)***	0.1695(0.097)*	2.7800(0.023)***
Constant	1.6448(0.069) ***	2.6524 (0.032)***	2.8341(0.040)***	4.1239(0.059)***
N	24,269	42,712	31,506	20,853
R	0,031	0,018	0,022	0,024

Robust standard errors omitted for brevity. *** p < 0.01, ** p < 0.05, * p < 0.10

To improve clarity, baseline regression results are presented for selected benchmark years representing different reform phases. Full year-by-year estimates and extended model specifications are available upon request.

Key within-year patterns of the explanatory variables are discussed below:

The dependent variable in this study is the natural logarithm of annual individual-level out-of-pocket (OOP) health expenditures (ln(OOP)). This transformation is used to reduce skewness and to allow coefficient interpretation in relative terms.

Age is measured as a continuous variable representing the individual’s age in years. It captures the effect of life-cycle health needs on healthcare spending.

Income is included in logarithmic form (ln(Income)) and represents annual disposable household income. The income variable used in the analysis corresponds to post-tax household income as reported in the TURKSTAT Household Budget Survey. This measure reflects the net income available to households after taxes and transfers. The logarithmic transformation allows for elasticity-based interpretation and reduces the influence of extreme.

Education is defined as an ordinal variable reflecting the highest level of education attained by the individual, with higher values corresponding to higher levels of educational attainment (e.g., primary, secondary, tertiary).

Gender is included as a binary variable (Female = 1), where male individuals constitute the reference category. This variable captures gender-based differences in healthcare utilization and expenditure patterns.

Marital Status is defined as a binary variable, where married individuals are coded as 1 and non-married individuals (including single, divorced, and widowed) are treated as the reference category.

Insurance Status is modeled using a set of dummy variables, with uninsured individuals as the reference category. The following categories are included:

• Pension Fund Insurance: Individuals covered under occupational or institutional pension fund schemes.

• Private Health Insurance: Individuals with supplementary private insurance coverage.

• General Health Insurance: Individuals covered under the national public health insurance system.

• Uninsured: Individuals without any form of health insurance coverage (reference category).

This classification allows for capturing potential differences in financial protection across different insurance types.

The constant term represents the baseline level of OOP expenditures when all independent variables are at their reference or zero values.

Socio-demographic determinants

As summarized in Table 2, the post-2017 expansion of PPP-based city hospitals, the scaling-up of digital health platforms, pharmaceutical cost-containment measures, and the broader post-2018 and post-pandemic policy environment jointly shaped the institutional context in which out-of-pocket expenditures evolved.

Reform periods and distributional effects

The income-based analysis reveals an inequitable distribution of the out-of-pocket (OOP) burden. Although the lowest income group (Q1) exhibits lower absolute levels of health expenditure, it bears the highest financial burden relative to income. While this study does not estimate population-level catastrophic health expenditure incidence, the observation that the out-of-pocket–to–income ratio for the lowest income group rises to the range of 8–12% in certain years suggests a heightened level of financial vulnerability, approaching commonly used WHO benchmark thresholds for catastrophic expenditure.

These patterns indicate that macroeconomic turbulence and reform dynamics after 2018 coincided with a weakening of financial protection mechanisms, particularly for lower-income households.

Discussion

This study shows how household out-of-pocket health expenditures (OOP) in Türkiye evolved over the period 2014–2023 and how the socio-demographic factors shaping these expenditures interacted with the period-specific dynamics of health reforms. The findings indicate that OOP expenditures are influenced not only by individual characteristics, but also strongly by economic conditions, health system reforms, and supply-side transformations. This evidence further confirms the necessity of adopting a political–economy–based approach to explain patterns in health expenditures. From a health equity perspective, the findings demonstrate that reforms expanded service availability but did not prevent the disproportionate concentration of financial risk among lower-income households. These distributional patterns provide the empirical basis for the equity-focused discussion that follows.

Effects of macroeconomic and reform dynamics

Health expenditures exhibited a continuous upward trend over the 2000–2023 period, with a marked acceleration after 2018. This pattern is closely associated with macroeconomic factors such as high inflation, exchange rate volatility, the rapid expansion of PPP city hospitals, and an increased shift toward private health services. The findings of this study indicate that OOP expenditures during this period were not solely driven by rising demand for health services but also reflected the financial implications of health reforms. These results are consistent with the literature emphasizing the heterogeneous cost effects of health reforms in middle-income countries [12, 24].

The widespread adoption of PPP city hospitals after 2017 exerted upward pressure on OOP expenditures through an expansion in high-technology–intensive services. Similarly, digital applications were not associated with a sustained reduction in observed OOP expenditure trends. This finding aligns with international evidence suggesting that digitalization does not necessarily lead to cost reductions in health systems [25].

Roles of individual determinants: continuity and period-specific differentiation

Regression analyses indicate that individual determinants such as age, education, gender, and marital status exerted consistent effects on out-of-pocket (OOP) expenditures across years. The pronounced influence of education and income levels suggests that health service utilization is shaped by social position and behavioral preferences. Higher OOP expenditures among women further indicate that health spending is closely linked to gender-based inequalities. These findings are consistent with the literature emphasizing the unequal distribution of health service utilization and financial burden [8, 31].

The inverse movement observed in income coefficients during the 2022–2023 period indicates that the post-pandemic context generated distinct dynamics. Temporary reductions in private hospital utilization among higher-income groups, the expansion of supplementary insurance coverage, and the concentration of deferred care among lower- and middle-income groups help explain these patterns. This finding demonstrates that the pandemic reshaped not only the supply of health services but also demand-side behaviors.

Reforms and inequalities: a political economy perspective

The findings of this study indicate that the effects of health reforms on financial protection in Türkiye have not been homogeneous across social groups. PPP investments and the expanding role of the private sector have altered the cost structure of technology-intensive services, generating a higher financial burden for lower-income groups. Similarly, restrictions in pharmaceutical policies, combined with increased use of over-the-counter products, have placed additional pressure on household budgets.

The distribution of the OOP burden across income groups appears inequitable. Although the lowest income group incurs lower absolute levels of expenditure, it bears the highest financial burden relative to income. In certain years, the proximity of this burden to the catastrophic expenditure threshold underscores the need to strengthen the inequality-reducing components of health reforms in Türkiye. This finding is consistent with discussions in the literature on welfare state transformation, particularly debates surrounding “hybrid models” that combine market mechanisms with state-led social protection [11, 17].

The out-of-pocket (OOP) health expenditures are examined not only on the basis of individual characteristics but also within the framework of the institutional structure and financing mechanisms of the health system. In this context, the political economy approach is integrated into the analytical process not merely as an interpretive framework, but also at the level of variable selection and model specification.

The variables included in the model—such as income, education, insurance type, and marital status—are considered not only as indicators of individual differences but also as proxies reflecting structural inequalities and resource allocation within the health system. In particular, the disaggregation of insurance types (general health insurance, private insurance, pension fund/bank-based insurance, and uninsured status) aims to capture differences in financial protection and the relative positions of actors within the system.

This approach is based on the assumption that OOP health expenditures are shaped not only by individual preferences or needs but also by policy choices, institutional arrangements, and the organization of healthcare service delivery. Accordingly, the findings are interpreted within the broader political economy context of the health system.

While the empirical model captures the overall effect of health system transformation through a post-2018 indicator, the reforms summarized in Table 2 differ substantially in their objectives, design, and mechanisms of influence. These reforms include pharmaceutical pricing regulations, insurance and financing adjustments, and changes in service delivery such as the expansion of public–private partnerships.

Although the data structure does not allow for the econometric isolation of each individual reform, their effects can be conceptually distinguished. For example, pharmaceutical policies may influence out-of-pocket expenditures through price controls and reimbursement mechanisms, while insurance-related reforms affect coverage depth and financial protection. Similarly, service delivery transformations may alter access patterns and cost structures across population groups.

From a political economy perspective, these reforms should not be interpreted as a uniform policy shift but rather as a set of structurally differentiated interventions that may produce heterogeneous distributional outcomes. Therefore, the observed effects reflect the combined and interacting influence of multiple reform trajectories rather than a single policy change.

Legal dimensions of out-of-pocket health expenditures (OOP)

Rising out-of-pocket health expenditures raise important legal and institutional concerns in Türkiye, where the right to health is constitutionally protected under Article 56. From a financial protection perspective, persistent increases in OOP spending—particularly among lower-income groups—highlight potential gaps in the state’s obligation to reduce financial barriers to access, as emphasized in international human rights and health financing frameworks (UN Committee on Economic, Social and Cultural Rights, [23], [24]).

The expansion of public–private partnership (PPP)–based city hospitals has intensified debates regarding fiscal sustainability, transparency, and long-term budgetary commitments. International evidence suggests that PPP-driven, technology-intensive investments may increase unit costs and indirectly shift financial pressure onto households, a pattern that has been observed across multiple health systems and is consistent with the post-2018 trends identified in this study [1, 4, 10, 28].

From a policy standpoint, these findings underscore the need for stronger regulatory instruments aimed at financial protection, including systematic monitoring of catastrophic health expenditures and the consideration of explicit limits on household out-of-pocket payments, as recommended in comparative health financing and universal health coverage literature [12, 26, 29].

Post-pandemic period: increasing pressures and new behavioral patterns

In the aftermath of the COVID-19 pandemic, the expanded role of public financing may have temporarily reduced out-of-pocket (OOP) expenditures in the short term; however, deferred care needs and subsequent utilization rebounds likely contributed to increasing demand during 2022–2023. This effect was particularly pronounced among individuals with chronic conditions and older age groups, contributing to the renewed rise in regression coefficients during this phase.

The pandemic reshaped both household health-seeking behavior and the relative weight of the private sector within the health system. While the expansion of digital health services facilitated access to care, it also intensified cost pressures through increased demand. This pattern is consistent with trends discussed in the literature as “post-pandemic cost acceleration” [18].

Comparative implications for middle-income countries

Türkiye’s experience demonstrates that health reforms in middle-income countries can generate complex outcomes in terms of financial sustainability and inequality. While instruments such as digitalization, PPP investments, and price-containment policies contribute to system modernization, they may also result in a partial shift of costs onto households. Therefore, policy design must consider not only service provision but also household financial vulnerability.

This study highlights the importance of integrating micro-level data analyses into policy evaluations when assessing the societal impacts of health reforms. Examining OOP expenditures through microdata in conjunction with period-specific reforms, enables the development of more sensitive and evidence-based policy recommendations.

One of the key strengths of this study lies in its analysis of out-of-pocket health expenditures (OOP) in Türkiye over an extended period using micro-level data, and in evaluating these trends alongside health reforms within a political economy framework. The large sample size of the TURKSTAT Household Budget Survey allows for the examination of changes in socio-demographic determinants over time, while multiple regression models enable a comparative assessment of the effects of individual factors across reform periods. The integration of reform timelines into the analysis represents a distinctive contribution that differentiates this study from much of the existing OOP literature.

Nevertheless, several limitations should be acknowledged. First, the microdata are based on self-reported expenditures, which may be subject to recall bias and measurement error. Second, although the HBS provides nationally representative data, the absence of detailed regional and geographic variables limits the ability to fully capture spatial inequalities in access to healthcare services. Third, unobserved factors such as regional price differences, variations in service availability, and private sector pricing strategies cannot be directly incorporated into the analysis.

A further limitation is the lack of detailed geographic identifiers (e.g., urban–rural or regional variables) in the publicly available TURKSTAT HBS microdata, which restricts the ability to analyze spatial inequalities in healthcare access and expenditure. As a result, geographic heterogeneity could not be incorporated into the empirical models. This constraint is acknowledged when interpreting the findings.

The inclusion of the impoverishment measure provides additional insight into the equity implications of OOP expenditures. While the observed impoverishment rates are relatively low, this finding should be interpreted with caution. Low impoverishment levels do not necessarily indicate strong financial protection but may instead reflect already constrained healthcare utilization among lower-income groups. In this context, households may avoid or delay care due to financial barriers, thereby limiting the extent to which OOP expenditures translate into measurable impoverishment.

This result is consistent with the broader literature, which suggests that low levels of measured impoverishment may coexist with unmet healthcare needs and access inequalities. Therefore, the findings should be interpreted alongside other indicators of financial burden.

In addition, the COVID-19 pandemic may have introduced temporary distortions in both healthcare utilization and reporting behavior, potentially affecting comparability across years. Finally, while the harmonization process ensured cross-year consistency, minor differences in survey implementation may still influence the measurement of certain variables.

These limitations suggest that the findings should be interpreted with caution, particularly when making cross-period comparisons or policy generalizations. Overall, the findings indicate that health reforms in Türkiye have expanded service capacity while leaving important equity challenges unresolved. From a political economy perspective, modernization strategies that rely heavily on private sector participation and digital expansion may unintentionally redistribute financial risks toward households, particularly those with lower incomes. These dynamics are not unique to Türkiye and offer broader lessons for middle-income countries pursuing universal health coverage under conditions of fiscal constraint and economic volatility. By drawing on a decade of nationally representative microdata, this study provides a systematic assessment of the long-term trajectory of out-of-pocket health expenditures in Türkiye. Integrating micro-level expenditure patterns with macro-level reform dynamics through a political economy lens, it contributes to equity-oriented debates on financial protection by demonstrating how post-pandemic health reforms have generated uneven distributional effects across income groups.

Conclusion

This study examined the evolution of out-of-pocket health expenditures in Türkiye over the period 2014–2023 and assessed how these expenditures were shaped by socio-demographic factors and the political economy of health reforms. The findings demonstrate that out-of-pocket spending is not solely driven by healthcare needs but is strongly influenced by macroeconomic conditions, structural reforms, and supply-side transformations. While health reforms expanded service capacity and improved access, they did not prevent the unequal distribution of financial risk across income groups. Lower-income households consistently faced a higher relative burden, highlighting persistent challenges in financial protection and health equity.

From a policy perspective, the results underscore the need to strengthen equity-oriented components of health reforms. High-cost investments such as PPP-based city hospitals require continuous assessment of their long-term budgetary and distributional impacts. Digital health applications, although effective in facilitating access, should be accompanied by regulatory mechanisms that enhance cost-effectiveness and prevent demand-driven cost escalation. Finally, the sustained vulnerability of lower-income households points to the importance of targeted financial protection measures, including exemptions from co-payments, need-based reimbursement schemes, and systematic monitoring of catastrophic health expenditures. These policy considerations are essential not only for Türkiye but also for other middle-income countries pursuing universal health coverage under conditions of fiscal constraint and economic volatility.

Abbreviations

CHE

Catastrophic health expenditure

GDP

Gross domestic product

HBS

Household Budget Survey

OECD

Organisation for Economic Co-operation and Development

OOP

Out-of-pocket (health expenditure)

PPP

Public–private partnership

SSI

Social Security Institution

TURKSTAT

Turkish Statistical Institute

UHC

Universal health coverage

WHO

World Health Organization

Author contributions

SC conceptualized the study, designed the methodology, conducted the data analysis, interpreted the results, and drafted the manuscript. AP, LÖ, and DŞ contributed to the interpretation of findings and critically revised the manuscript for important intellectual content. All authors read and approved the final manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The data that support the findings of this study are available from the Turkish Statistical Institute (TURKSTAT) under license. Restrictions apply to the availability of these data, which were used under permission for the current study and are not publicly available. The data are, however, available from the corresponding author upon reasonable request and with permission of TURKSTAT.

Declarations

Ethics approval and consent to participate

This study is based exclusively on anonymized and publicly available secondary microdata from the Turkish Household Budget Survey provided by the Turkish Statistical Institute (TURKSTAT). The dataset does not contain any identifiable personal information. In accordance with national regulations and institutional policies, ethical approval and informed consent were not required.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.