![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
Does foreign aid strengthen or weaken post-conflict states? We examine the effects of aid on tax collection after civil war, an important dimension of state effectiveness. While the literature emphasizes aid’s perverse effects, the relationship between aid dependence and the growth of tax collection is unclear. We argue that the impact of aid reflects its political utility for ruling elites in consolidating their authority after civil war. While dominant parties subvert tax strengthening reforms to solidify their political base, elites in more fractionalized settings rely on external political backing to manage internal challenges to their authority, and are more likely to comply with donor conditions. We test this argument through a Latent Curve Analysis of tax collection rate growth in post-civil war countries from 1978 to 2012. We find that aid is associated with slower growth in tax collection in dominant party settings, and more rapid tax growth in politically fractionalized settings. The findings highlight the need for attention to internal political dynamics to explain aid effectiveness after civil war, and point to opportunities to strengthen institutions in some post-conflict countries.
Acknowledgments
We thank Naazneen Barma, Susanna Campbell, Desha Girod, Daniel Tirone, Christoph Zürcher, the editors and anonymous reviewers for their very helpful feedback, and Michael Jablonski, Clark DeMas, Dariga Abilova, and Jason Levitt for valuable research assistance.
Supplementary Material
The supplemental data for this article can be accessed here.
Correction Statement
This article has been republished with minor changes. These changes do not impact the academic content of the article.
Notes
1 Tax collection also reflects policy preferences regarding tax rates, but the need to increase tax collection is common after civil war.
2 These trajectories and the data points from which they are calculated are shown for each of the peace spells we analyze in the supplementary appendix.
3 Not all tax exemptions reflect these dynamics, the key distinction is whether these are made outside of the formal policy process, typically by the executive branch.
4 Revenue increased after 2002, once the ruling party was firmly in power, and mostly due to a new VAT tax as income and business taxes stagnated.
5 Minor wars are less likely to be followed by reconfiguration of power, large influxes of aid, and dependence on donors to consolidate authority. While they may involve similar political dynamics, they are beyond the scope of our argument and we leave this investigation to further research. We follow the conventional definition of major civil war as an armed conflict between at least two organized parties within a state that results in at least 1,000 battle-related deaths.
6 Since we focus on state-level changes, we do not count periods after simultaneous civil wars in a single country as separate peace spells.
7 The sample is limited to post-1978 by data availability for the main dependent variable, tax revenue.
8 The distribution of slopes is displayed in the Supplementary Appendix, Figure A2.
9 We employ a two-year timeframe for the independent variables to mitigate endogeneity concerns. These concerns are further discussed below.
10 Although not all donors are interested in tax policy, the vast majority of post-conflict aid-dependent countries have adopted IMF programs with conditions on revenue generation and/or received donor assistance related to tax reform.
11 Unlike Greene (Citation2010), our measure of party dominance does not include parties’ longevity over time.
12 Data were collected from the Inter-Parliamentary Union’s PARLINE Database on National Parliament, supplemented with data from national elections commission websites and secondary sources. Seat share is calculated for the lower house in bicameral systems.
13 According to the Polity IV index, Liberia had a score of 5 and Mozambique had a score of 4.
14 Data from the WDI.
15 Data from the WDI.
16 Data from Fearon Citation2003.
17 Data from USAID Citation2010.
18 We recognize that this not a complete fix for potential endogeneity, but contend that our modeling approach is attuned to this issue. Since our model investigates the effect of aid dependence and party dominance in the first two peace years on the slope of the tax collection trajectory for the entire peace spell, we minimize the possibility that changes in tax collection are the root cause of our observed effects. Of course, we cannot rule out the influence of unobserved correlates, but in a series of robustness checks, we tested a wide range of models with a variety of control variables, and using a variety of time horizons for the independent variables, with consistent results. These tests are discussed below and the results are included in the appendix.
19 We used the average score over the first 2 years of the peace spell for our main independent variables. For control variables, we used averages over the entire peace spell, as none of the controls are highly variable over time.
20 The tax rate trajectory for each of these cases is displayed in the Supplementary Appendix, Figure A1.
21 Some of the reforms were later criticized for adverse effects on peace. See Paris Citation2004.
22 The estimated effect of party dominance in aid-dependent countries is −1.52 percentage points, which corresponds to a predicted difference between Laos and Liberia, for example, of 12.2% in the tax collection rate over a decade. Both countries are aid dependent but Laos’s dominance score is 1, compared to Liberia’s score of 0.2.
23 These base models address the potential problem of overfitting, which can arise when there are too few observations and too many variables in a model.
24 Party fractionalization is defined as the number of parties in the legislature weighted by vote share. See Rae Citation1971.