Electoral Incentives, Distributive Politics, and Tax Evasion in Pakista

How do electoral incentives shape officeholding politicians' decisions and behavior? Though political science has long analyzed how governments distribute development resources among different parties and districts, the focus has primarily been on stable political systems. The first two chapters of this dissertation extend our understanding of how political considerations affect resource distribution by incumbent parties in politically competitive but institutionally unstable democracies. The third chapter shifts attention to how electoral pressures condition legislators' own behavior in terms of tax payments in response to increased information about tax evasion to citizens. Empirically, the dissertation focuses on Pakistan, which is an unstable democracy but one that has nonetheless experienced frequent elections with high competition. All three chapters use new data that I collected myself doing fieldwork in Pakistan.;In the first chapter, I argue that the specter of regime instability shortens politicians' time horizons when they are in power in countries that have faced frequent regime alternation. The consequent expectation that democracy, when in place, is unlikely to last in the long term incentivizes higher levels of resource distortion than in stable systems because it lowers the expected future costs of punishment for such behavior. Thus, in unstable democracies, we should see legislators from the ruling party having disproportionately high access to development resources compared to opposition legislators. Using the first comprehensive dataset on two decades of federal development resource allocation and national elections in Pakistan, and implementing a regression discontinuity design, I show that the ruling party greatly distorts access to resources in order to benefit its own party members and punish opposition legislators. I also present evidence that this distortion is driven by the treat of regime instability rather than by legislator quality, economic differences between districts, the timing of elections, or ideological differences between the main political parties.;The first chapter's findings lead to the question of whether the lower access to resources for opposition legislators is driven by all opposition legislators being discriminated against equally or whether other factors influence the distributive pattern within opposition districts. I argue in the second chapter that high electoral competition and a patronage-heavy political environment act as a constraint on the ruling party such that it will not discriminate against the strongest opposition legislators when distributing resources. A simple formalization of the main mechanism shows that, under such conditions, the difference in resource access for ruling party versus opposition legislators is driven by high discrimination against opposition swing legislators rather than opposition core legislators who, in fact, are treated similarly to the ruling party's own districts. Using district- and administration-level fixed effects estimators along with the same development fund data introduced in the first chapter, I find strong empirical support for the proposed hypotheses.;In the third chapter, I address how an increase in transparency --- in this case, in the form of information to citizens --- affects legislators' subsequent behavior, and how this response is conditioned by the different electoral pressures that different groups of legislators face. Exploiting an unforeseen decision by the Pakistani government to publish and publicly share tax payment records of all legislators several months after tax returns had already been filed for the previous year, I use this exogenous 'information shock' to measure its effect on legislators' subsequent tax payments. Furthermore, I show that the pressure to respond to such information release is higher for directly elected legislators, and for legislators elected in competitive races. I introduce new data on politicians' asset ownership and federal tax payments, and use it within a difference-in-differences research design to show strong evidence of heterogeneous responses from these groups of legislators, even after controlling for other relevant differences.;Together, the dissertation contributes to several important literatures. The first two chapters extend the existing distributive politics literature to focus on countries with unstable regimes by addressing how the expectation of instability affects political decision making. The arguments and findings also contrast starkly with the analogous literature on developing countries that, instead, finds that the intensity of punishment increases as opposition support increases. The second chapter also has potential implications for the broader literature on development. The highest variance in access to resources emerges in the most competitive districts, while strongholds controlled by either party have relatively stable access to resources. In the long term, this pattern could lead to distorted development outcomes. Chapter 3 speaks to the literatures on corruption and on accountability of elected officials to voters. How increased transparency affects corrupt behavior is of great interest to scholars but is difficult to observe and accurately measure, especially outside of experimental settings. Furthermore, by analyzing the subsequent behavior of legislators conditional on the intensity of electoral pressures they face, the findings also add to our understanding of the effectiveness of electoral incentives as accountability mechanisms for political representatives.