When Faced with Income and Asset Shocks, Do Poor Rural Households in Vietnam Smooth Food Consumption or Assets?

Abstract

Vietnamese rural households are exposed to severe covariate and idiosyncratic shocks. However, these households are remarkably resilient and have steadily increased real income and consumption over the survey period 2006 to 2012. To explain household strategies to cope with shocks we test three theoretical models: the Permanent Income Hypothesis (PIH), Complete Market Hypothesis (CMH), and Asset Smoothing Theory (AST). There is support for AST as households smooth productive assets rather than consumption; and for CMH as households smooth consumption against idiosyncratic, but not covariate shocks. There is no support for PIH.

Acknowledgements

The authors would like to thank DANIDA for making the data available for this study and Australia Awards Scholarship for scholarship funding. Data and Stata codes will be made available on request.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplementary Materials are available for this article which can be accessed via the online version of this journal available at https://doi.org/10.1080/00220388.2018.1528350

Notes

1. We find the model free from omitted variable bias. We test for this using a Ramsey regression specification-error test for omitted variables applied to the income Equation (1). We cannot reject the Null Hypothesis of no omitted variables (F-stat = 1.58, P-value = 0.19).

2. The quasi-experimental approach will not capture other transitory and permanent effects of unobserved shocks, which are possible using the statistical approach (Blundell, Pistaferri, & Preston, 2008). However, use of observable shocks will allow the simultaneous test of CMH and the distinction between asset shocks and income shocks.

3. Early tests of CMH tend to use group average consumption or average income to proxy for the covariate shock and individual income for the idiosyncratic shock (Kurosaki, 2006; Skoufias, 2003; Townsend, 1994; Wong & Godoy, 2003). Fafchamps and Lund (2003) argued against construction of shocks from income data and in favour of observable shocks because they are less subject to measurement error and allow both consumption and income shocks analysis. Schulhofer-Wohl (2011) showed that construction of shocks from income in the early CMH tests can cause bias because of the heterogeneity in risk preference. Observable shocks were used by Gertler and Grubert (2002) and recently by Gaurav (2015).

4. We are grateful to a referee for pointing this out.

5. A version of AST that uses the growth rates in productive assets balances leads to stronger AST support. For 1 per cent of income losses caused by floods and crop diseases, productive asset grows significantly at the positive rates of 4 per cent and 6 per cent respectively.

6. AST test using productive asset growth rates shows that only agricultural households significantly increase their investments in productive assets when income is reduced by crop diseases and floods.