- Main
Essays on Local Public Finance and Education in Paraguay
- Meza Arrua, Sara Maria Ines
- Advisor(s): Carter, Michael
Abstract
This dissertation comprises three essays that examine, on the one hand, how local governments respond to additional resources from the central government and, on the other hand, whether an innovative pedagogical methodology improves student outcomes. The three essays focus on the context of Paraguay, where municipal fiscal capacity remains limited and municipal budgets are highly dependent on central government transfers. At the same time, educational challenges persist in this context, making the evaluation of pedagogical innovations particularly relevant. The first chapter studies the effects of intergovernmental transfers on local public finance outcomes. In Paraguay, municipalities receive transfers that come mainly from the exploitation of two binational hydroelectric dams. These transfers are partially earmarked for investment purposes. To estimate the causal effects of these hydroelectric dam revenues, I leverage a legally predefined allocation formula used to distribute these revenues that combines an equal-share component with a population-based component within a shift-share instrumental variables design. Results show that transfers generate an expansion of local fiscal activity. On the revenue side, higher transfers increase local tax collection, with positive effects on both property taxes and other locally administered taxes. On the expenditure side, municipalities exhibit a strong flypaper-type response, with increases in total expenditures close to or above one-for-one, in accordance with previous studies on public spending responses to additional transfers. While a significant share of the additional resources is allocated to capital expenditure, consistent with partial earmarking rules, municipalities also increase current expenditure. I then examine the mechanisms driving these results. I find that increases in other taxes are concentrated in fees and charges linked to local service provision, while property tax effort remains unaffected. Regarding expenditures, the expansion is driven by both investment-related categories and current expenditure, particularly in personnel and operational categories. These patterns suggest that municipalities respond to additional resources by scaling up existing fiscal instruments and administrative structures, rather than undertaking structural improvements in tax enforcement or public finance management. The second chapter builds directly on the findings of Chapter 1 and investigates whether all municipalities respond similarly to additional transfers. Paraguay has experienced sustained economic growth since the mid-2000s, but this process has been uneven across municipalities, leading to disparities in local economic conditions and institutional capacity. These differences imply that similar increases in transfers may translate into heterogeneous fiscal responses. To examine this possibility, I first develop a political economy model of local fiscal behavior in which municipalities differ in their effective public goods technology and the political incentives they face. In this model, municipalities operate in a low- or high-spending regime determined by a public spending threshold. Preferences and composition of informed voters then decide the likelihood of crossing the threshold, while the response of local taxation to transfers depends on the marginal productivity of public spending and the returns to scale of public goods. Next, I test the model predictions empirically by implementing two complementary approaches: (i) I estimate interaction models that allow the causal effect of transfers to vary with predetermined municipal characteristics within the same shift–share instrumental variables framework used in Chapter 1, and (ii) I estimate mixed-effects models that recover municipality-specific responses to predicted transfers, allowing for unobserved heterogeneity in fiscal behavior. Results demonstrate that municipalities with stronger economic bases, greater urbanization, better infrastructure, and more diversified economic structures tend to exhibit amplified fiscal responses, with increases in both local taxes and public spending. In contrast, more rural and less developed municipalities tend to display weaker fiscal responses and patterns consistent with tax crowd-out. The mixed-effects estimates further show that municipalities are distributed across distinct response regimes, including cases of symmetric high or low responses, and asymmetric adjustments across the two fiscal margins, local taxes and public spending. These patterns suggest that the effectiveness of transfers depends critically on local conditions that shape both the productivity of public spending and the incentives for revenue mobilization. The third chapter, co-authored with Claudio Cappello and José Molinas Vega, analyzes the effects of an arts-based and active learning methodology, called “Pre-Texts,” on students' socioemotional and academic outcomes. The methodology combines reading, artistic production, collective discussion, and creative interpretation of texts to promote student engagement with learning activities. To evaluate the effects of this methodology, we run a randomized controlled trial to sixth-grade students in public schools in collaboration with the Paraguayan Ministry of Education and Science and the local partner Instituto Desarrollo. Results indicate moderate but suggestive effects concentrated primarily in reading-related dimensions. On average, the intervention increases reading enjoyment and reading motivation. We also find evidence of positive effects on academic achievement in Spanish, measured as the probability of achieving the highest score, while effects on Mathematics and grade repetition are limited. Regarding heterogeneous effects, reading outcomes tend to exhibit somewhat larger positive effects among boys and students attending morning-shift schools, while effects for other subgroups are generally smaller and not statistically significant. Taken together, findings suggest that the Pre-Texts methodology may operate primarily through mechanisms related to students’ engagement with reading and classroom participation rather than through short-run transformations in socioemotional well-being. This interpretation is consistent with the structure of the intervention itself, which places collective reading and participation at the center of classroom activities.