Declining Poverty and Inequality in Turkey: The Effect of Social Assistance and Home Ownership

ABSTRACT

Social assistance has become prominent in combating poverty in developing countries, and has also contributed to the popularity and election success of governments implementing it. In this paper, I employ household surveys and investigate the effect of social assistance on poverty and income inequality in Turkey. I also review the recent literature on poverty, as well as different components of social protection spending: education, health, pensions and housing. In the empirical analysis, I show that pensions still constitute the bulk of public transfers to households. Moreover, home ownership ameliorates poverty and inequality for Turkey. Despite its modest amounts, social assistance reduces poverty and its marginal effect on income inequality is larger than other income sources. These findings suggest that increases in social assistance budgets should accompany other policy measures in combating poverty and inequality.

KEYWORDS:

• Social protection
• social assistance
• poverty
• income inequality
• housing
• Turkey

Acknowledgments

I am very grateful to the anonymous reviewers and Editors Susannah Verney and Anna Bosco at South European Society and Politics for their constructive criticisms and suggestions that significantly improved the original draft. I also thank Doruk Cengiz, Hasan Cömert, Yasemin Dildar, Çağla Diner, Tim Dorlach, Emine Tahsin, Zhun Xu, Seyit Cilasun and Alper Yağcı for their detailed criticisms, as well as the participants of the Turkish International Political Economy 2016 Workshop for their comments and suggestions for earlier drafts. All the remaining errors are mine.

Disclosure statement

No potential conflict of interest was reported by the author.

Supplementary material

The supplementary material for this paper is available online at https://doi.org/10.1080/13608746.2018.1548120

Notes

1. The study by Gidengil and Karakoç (2016) is concerned with voter intentions before June 2011, and their survey questions on social policy bunch together access to healthcare, education, and non-contributory social assistance but not pensions. In other words, their survey does not separately investigate the impact of each dimension of social policy.

2. Official statistics on social assistance show a marked increase in 2012 which is due to reclassification of healthcare spending for the very poor as social assistance. In the 1990s and early 2000s, spending on the Green Card (health services for people with no social security) was part of the Ministry of Health Budget. With the transition to a General Health Insurance Scheme, the Ministry of Family and Social Policy (MFSP) became responsible for paying the health insurance premiums of the poorest households. In accordance with this re-organisation, Green Card funds are re-classified as social assistance. In other words, neither the poorest households nor the MFSP ever touch this largest chunk of its budget (around 40 per cent every year). What used to be healthcare spending was suddenly reclassified as social assistance, and on paper overall social assistance increased by 0.4–0.5 per cent of GDP.

3. Legislation that influences the bargaining power of workers vis-à-vis employers (such as minimum wage and ease of unionisation) also can influence the income distribution.

4. Gini coefficients for land inequality in 1960 were: Bolivia 0.768, Argentina 0.814, Brazil 0.787, Chile 0.865, Colombia 0.805, Ecuador 0.804, Paraguay 0.863, Uruguay 0.791 versus Turkey 0.608 versus South Korea 0.364 and Japan 0.398.

5. Detailed comparisons of government revenue and expenditure on different dimensions of social protection spending of selected Latin American countries, Turkey and South Korea employing Worldbank Databank (2017) is available from the author upon request.

6. Buğra (2018, pp. 327–328) provides an overview of the most recent picture of in-kind public care services.

7. Başlevent (2016) also concludes that social assistance is relatively better captured in the SILC dataset for 2012. I could not obtain comparable estimates for households’ rental income and there is no other source for estimating private transfers between households.

8. The poverty gap measure answers the question: if we can identify and target poor individuals perfectly, what percentage of the poverty line, on average, will suffice to raise every individual to the level of the poverty line? The squared poverty gap is similar but gives greater weight to the poorest of the poor.

9. Turkstat stopped publishing an absolute poverty threshold in 2010; I simply inflated the 2009 threshold using the consumer price index. Turkstat (2017) official poverty line (TL) varied with respect to household size (household size in parenthesis): 5,554 (1); 8,339 (2); 10,560 (3); 12,464 (4); 14,262 (5); 15,863 (6); 17,419 (7); 18,900 (8); 20,184 (9); 21,498 (≥ 10).

10. This overall stability hides the declining homeownership trends in Istanbul and Ankara.

11. Tables corollary to Table 2 and separately for regular and irregular assistance are available from the author upon request.

Additional information

Funding

This work was supported by the Kadir Has University; [2017-BAP-03].

Notes on contributors

Hasan Tekgüç

Hasan Tekgüç is an Associate Professor of Economics at Kadir Has University. His research interests include poverty, gender issues in labour markets, access to education and migration.