foreign aid and domestic tax
DESCRIPTION
Foreign Aid and Domestic Tax: Multiple Sources, One Conclusion -Presentation by Paul Clist (U. East Anglia). The relation between taxes and foreign aid in developing countries has been vastly investigated in the literature. Indeed, in the short term, aid inflows may be seen as a substitute for tax revenues in the eyes of recipient countries, thus lowering the incentive of the latter at increasing their revenues from taxation. However, in the long run, relying on foreign aid with no or marginal taxing of the people might affect governance of the countries, as it breaks the social contract between the citizen and the state. One common finding in the literature is the negative impact of grant type of aid on recipients’ government revenues. This paper therefore sought to replicate existing studies, adding new inputs in order to see whether this assertion would remain true. The paper by Paul Clist failed at replicating Benedek et al. (2012), which concluded that there is a negative association between domestic tax mobilization and grants, though this relationship appears to have weakened as the original dataset exhibits some discrepancies. Clist and Morrissey (2009) extended Gupta et al. (2004) study over the period 1970-2005 (against 1970-2000). Although they found a significant negative impact of grant aid on government revenue, this effect happened to be driven by the period 1970-1985, when the IMF data was at its worst. When focusing on the period 1986-2005, results are reversed with a robust positive relationship between grant aid and tax revenues. In terms of limitation, one of the main issues stem from the fact that most of the previous studies were conducted using data from different sources, where key variables had different definitions. Endogeneity is also a major concern as poor tax collection performance and increased aid grant can simultaneously respond to the same shock such as natural disasters. Four strategies to cope with those problems are established in the paper, including one year and two year lagged aid variables, multiple indicators with multiple causes (MIMIC) models and the use of the new ICTD GRD. In most of the cases, the statement that aid grant has a negative impact on tax revenue is rejected and, as the new data shows, there is no negative effect of aid grants on tax revenue at all. The conclusion that can be drawn from this is that the negative relationship between tax revenue and grants is driven by bad quality data, making the ICTD GRD not only better, but also much needed.TRANSCRIPT
Foreign Aid and Domestic Tax: Multiple Sources, One Conclusion
Paul Clist 09/09/2014
Why do we care about Tax? 1: short term, crowding out of aid revenue 2: long-term, corrosion of governance
2/7
Gupta et al (2003) “… while concessional loans are associated with higher levels of domestic revenue mobilization, the opposite is true of grants”
Loan 0.011***
(3.05)
(Loan)² -0.0001
(-0.99)
Grant -0.016***
(-5.00)
(Grant)² 0.0004***
(5.07)
1970-2000, 107 countries OLS with fixed or random effects RHS also includes agriculture, industry, GDP pc, Trade openess
3/7
Clist & Morrissey (2009) Extend data by five years 1970-2005 ‘trade’ →exports and imports (one-year lag on aid)
Loan 1.25***
(3.57)
(Loan)² -0.003*
(-1.78)
Grant -1.18***
(-5.42)
(Grant)² 0.0003***
(1.49)
This effect seems to be driven by the period 1970-1985 (worst data) All types of aid is positive over 1985-2005
Loan 1.69***
(3.90)
(Loan)² 0.003
(1.40)
Grant 0.80***
(2.80)
(Grant)² 0.0001
(0.52)
Loan 1.25***
(3.57)
(Loan)² -0.003*
(-1.78)
Grant -1.18***
(-5.42)
(Grant)² 0.0003***
(1.49)
4/7
Benedek et al (2012) 118 Countries, 1980-2009, up to 2589 obs. OLS and GMM
“Overall... a negative association ... but this relationship appears to have weakened ...”
Carter (2010, 2013) critique: • Endogeneity not dealt with • GMM really problematic
5/7
In Clist (2014) I can't replicate Benedek et al Sample discrepancy e.g. 6 countries apparently in the Benedek et al dataset are not in the provided dataset. Large discrepancies between reported variables and reconstructed data e.g. trade openness mean of 18.75 (reported) 54 (mine).
Larger problem: The dependent variable is actually taken from 7 sources Average yearly difference for same country with same source: 1.3%, for different source: 4.9% One source measures all government revenue
6/7
What to do? • Source coefficients/slopes/separate regressions → this is a problem • MIMIC model → no negative effect from aid grants • New data → there is no negative effect
The only time I find a negative effect is contemporaneously for aid grants – ignoring endogeneity. Reverse causality seems a more plausible explanation. New data is much needed.
7/7
• Gupta S, Clements B, Pivovarsky A, Tiongson E. 2004. Foreign aid and revenue response: does the composition of aid matter? Chapter 14. In Helping Countries Develop: The Role of Fiscal Policy, Gupta S, Clements B, Inchauste G (eds). International Monetary Fund: Washington DC; pp. 385–406.
• Clist, P., and O. Morrissey, 2011, “Aid and Tax Revenue: Signs of a Positive Effect since the 1980s,” Journal of International Development, 23, pp. 165–80.
• Carter, P., 2013, Does Foreign Aid Displace Domestic Taxation?. Journal of Globalization and Development, 4(1) pp. 1-47.
• Benedek, D., Crivelli, E., Gupta, S., and Muthoora, P. (2012) Foreign Aid and Revenue: Still a Crowding Out Effect?, IMF Working Paper, 186.
• Clist, P. (2014) Foreign Aid and Domestic Tax: Multiple Sources, One Conclusion, ICTD Working Paper