The impact of fiscal decentralization on health outcome in Vietnam: a focus on the role of public service delivery and corruption

ABSTRACT

This study examines the applicability of fiscal decentralization theory, originally developed in Western countries, to Vietnam at the provincial level. The aim is to explore the indirect impact of public service delivery on the relationship between fiscal decentralization and health outcomes, with corruption as a moderator. The study employs the conditional process analysis. The findings reveal that the multi-dimensional nature of fiscal decentralization results in different statistically significant effects on health outcomes. While fiscal autonomy does not guarantee an impact on health outcomes, revenue-fiscal decentralization positively improves health outcomes. These effects are intensified by more significant control over corruption, suggesting that limiting corruption is crucial to ensuring the benefits of fiscal decentralization. Moreover, our study shows that public service delivery does not mediate the effect of fiscal decentralization on health outcomes. Overall, our findings suggest that applying Western theories of fiscal decentralization to the context of Vietnam is challenging. The research recommends reforming the fiscal transfer mechanism and strengthening control over corruption in public services.

KEYWORDS:

• Fiscal decentralization
• health outcome
• public service delivery
• Vietnam local government

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/17516234.2023.2284440.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data source

Vietnam General Statistic Office: https://www.gso.gov.vn/en/statistical-data/

The Ministry of Finance: https://ckns.mof.gov.vn/eng/SitePages/Home.aspx

PAPI Index: https://papi.org.vn/eng/du-lieu-papi/

Notes

1. The Heteroskedasticity Consistent Covariance Matrix Estimators (HCCME) method varies depending on the approach to adjust the covariance structure of the error term. According to MacKinnon and White (1985), HC3, called the “jackknife, performed the best (SAS HELP). In addition, Long and Ervin (2000) compared the HCCME method using Monte Carlo simulation and found that HC3 is most appropriate in the case that the sample size is below 250.

2. Number of bootstraps for confidence intervals was set to 10,000 and the seed number was set to 12,345.

Additional information

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes on contributors

Hung Dao

Hung Dao is a Ph.D. candidate at the Graduate School of Public Administration at Seoul National University (South Korea). He is interested in policy analysis, policy evaluation, decentralization, and e-Government.

Minjun Hong

Minjun Hong is a Ph.D. candidate at the Graduate School of Public Administration, Seoul National University, South Korea. His research interests include policy analysis, cost-benefit analysis, and local government services.