The impact of universal health coverage on households’ consumption and savings in Thailand

ABSTRACT

This paper studies the impact of Thailand's Universal Health Coverage Scheme (UCS) on households' consumption and savings by using a synthetic panel data approach. Using difference-in-differences estimation, it finds that, in the short run, the UCS had little or no impact on either savings or consumption expenditures. In the long run, the UCS still had no effect on savings, but there is some evidence of an increase in consumption. These effects are generally consistent with economic theory. The provision of healthcare coverage at little or no cost to previously uninsured households has an income effect that will increase both savings and consumption and a risk reduction effect that will reduce precautionary savings and thus increase consumption. These two effects on savings are of opposite sign and appear to cancel each other out, while both effects on consumption are positive and so appear to increase consumption, at least in the long run.

KEYWORDS:

• Universal coverage
• health insurance
• precautionary savings
• synthetic panel

Acknowledgments

The authors would like to thank the National Statistical Office of Thailand for providing data used in this study, and the participants in various conferences and meetings in which earlier versions of this paper were presented.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The Medical Welfare Scheme (MWS) later became the Low-income Card Scheme (LICS), but the two terms will be used interchangeably throughout this paper.

2. The CSMBS provides benefits to government employees, their dependents (including spouse, parents and up to three children below 20 years old) and public sector retirees. The Social Security Scheme (SSS) is a compulsory scheme for employees, but not their dependents, of private enterprises with at least 20 employees.

3. The VHCS required each enrolee to purchase a health card that cost 500 baht per year.

4. The UCS is also known as the 30-baht scheme or the Gold Card (GC) scheme. The former refers to the 30-baht copayment, and the latter refers to the health cards issued by this programme. Regular Gold Cards were issued to beneficiaries required to pay the 30-baht copayment, and Gold Cards with exemption (GCE) were issued to those exempted from copayments. This paper will use the term Universal Health Coverage Scheme (UCS) instead of 30-baht scheme, and it includes both GC and GCE. (As explained below, 30-baht payments were later discontinued.)

5. Funding for the SSS is from three sources, the employees, the employers and the government.

6. These figures are from official government reports; the survey data below show a small group of people who report not being insured; they are covered by UCS but have not yet used it.

7. The UCS's sustainability is a serious problem. Many hospitals have budget deficits, but others do not. However, the financing of the UCS is beyond the scope of this study.

8. Data from the years after 2007 are not used because the differentiation between UCP and UCE is available only until 2007, which is the year when the copayment was eliminated.

9. Thailand has 76 provinces; nearby provinces with similar characteristics are grouped according to the Ministry of the Interior classification scheme, which yields 19 groups of provinces.

10. The only exceptions are the Voluntary Health Card Scheme (VHCS) in the pre-2001 period, which covered all household members, and the CSMBS in all years, which covers not only government workers but also their dependents (spouse, parents and children up to age 20).

11. NTASST cannot be used to compare savings between 2001 and 2007 because the 2007 SES data on assets and liabilities is limited to a set of categorical variables.

12. Durable expenditures can also be defined as a component of savings since they often involve instalment plans; when spending more on durables, individuals tend to reduce future savings.

13. Theoretically, anyone who remains uninsured after the UCS implementation is entitled to UCS coverage automatically. However, some people may report that they are still uninsured (as found in the 2004 and 2007 HWS data) after the UCS was implemented. They could be individuals not covered by the UCS due to legal constraints (e.g. illegal immigrants), or people who currently have no need to seek medical care and hence have not (yet) obtained a UC card. The 2004 and 2007 HWS data indicate that the people who remained uninsured are mostly the latter. Given this ambiguity, and the small number of uninsured in the 2004 and 2007 HWS data, those who report being uninsured after the UCS implementation are excluded from the analysis. Note that a robustness check that includes this group as UCS beneficiaries yields very similar estimates.

14. The empirical analysis uses information on health insurance from the 2001 HWS to predict whether the UCP beneficiaries in 2004 and 2007 were uninsured or covered by VHCS in 2001. This is done by using a linear probability model, as explained in the data section.

15. This choice for the comparison group is consistent with that of Ghislandi, Manachotphong, and Perego (2015).

16. The Socio-Economic Survey (SES) has a variable for households’ socio-economic class. It is computed using the household's land ownership and the work status and occupations of household members, and the economic sectors in which they work. This variable has seven categories: (1) farm operator (own land), (2) farm operator (rent land), (3) entrepreneurs, trade and industry, (4) professional, technical and managerial, (5) labourers, (6) other employees and (7) economically inactive.

Additional information

Notes on contributors

Phatta Kirdruang

Phatta Kirdruang is an assistant professor at the Faculty of Economics, Thammasat University, Thailand. She received her PhD degree in applied economics from the University of Minnesota. Her research interests include health economics and development economics.

Paul Glewwe

Paul Glewwe is a professor in the Department of Applied Economics at the University of Minnesota. He received his PhD in economics from Stanford University. His research interests include development economics and economics of education.