Growth versus equity: the effects of centralized fiscal transfers on Chinese provinces

ABSTRACT

Centralized fiscal transfers to the provinces are the principal means of reducing income inequalities in China. The coastal provinces are the main transfer donors to the poorer provinces. We evaluate if the equity objective sacrifices aggregate growth using four panel vector auto-regressions (PVARs) for four geographical regions of Chinese provinces. The PVARs are estimated for centralized transfers, government spending, taxes and output from 1994 to 2018, whether or not transfers sacrifice growth depends on geography. Transfers reduce inequality and generate growth in the middle provinces; reduce inequality but sacrifice growth in the western provinces; and worsen inequality and sacrifice growth in the northeast provinces.

KEYWORDS:

• regional growth
• regional inequalities
• fiscal multipliers
• fiscal transfers
• decentralization
• China

JEL:

• C32
• E62
• H61
• R11

ACKNOWLEDGEMENT

We gratefully acknowledge the helpful comments of the editor and three anonymous referees, and of the participants at the Chinese Economists Society annual conference 2021, the Cardiff China Shadow Bank Workshop 2021, the joint conference of Peking University and University of Nottingham Ningbo 2021, and the 8th International Conference on the Chinese economy, Tsinghua University, December 2021. We especially thank Ron Smith (Birkbeck London) and Steven Dellaportas (Xi’an Jiaotong–Liverpool) for their helpful comments and advice. All remaining errors are ours alone.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. Wang et al. (2020) also use a DSGE framework to explain the responses of government investment, social investment and household consumption across regions to an increase in centralized transfers established in a single PVAR.

2. The trade-off between bias and variance is central in econometrics. Adding parameters reduces bias but increases variance. The model selection criteria consider both fit and parsimony (having fewer parameters). We use the Schwarz BIC, which selects a more parsimonious model than the Akaike information criteria (AIC) (Smith & Fuertes, 2020).

3. Xinjiang and Tibet have negative values of IGE in some years so we exclude these two provinces. Negative values of IGE imply that the centralized transfers exceed the total provincial spending (GOV) that could be saved and spent in the subsequent years.

4. See Rudebusch (2005) for US monetary policy rules; and Smith (2009) for the European Monetary Union.

5. We thank Yong Wang from Peking University for making this point at the joint UNNC–Peking conference. The negative multipliers could also be due to a potential endogeneity issue. If the government anticipates an economic decline, it would increase transfers, inducing a negative coefficient. Output could have been much worse had it not been for the transfers, but the model does not have this information, thus output appears less negative from the model than what it would have been, so it is reflecting the endogenous response than the causal effect of transfer on the economy. The estimate could be a mixture of both effects: a displacement of the non-government sector and the endogenous response of the central government in anticipating an economic decline in the northeast.

6. We thank an anonymous referee for raising this point.

7. Shan et al. (2020) highlight issues in estimating potential output in China when estimating a monetary Taylor rule.

Additional information

Funding

This work was supported by the Economic and Social Research Council (ESRC)-Newton fund [grant number ES/P004199/1]; the National Natural Science Foundation of China [grant number 71661137005]; and the National Social Science Fund of China [grant number 22BJY032].