Ana Paula Franco

This paper studies how firm–worker agreements to underreport wages affect tax policy design. In most tax systems, firms act as third-party reporters to limit evasion, but this mechanism weakens when employers pay part of workers’ wages off the books—what I refer to here as payments under the table (PUT).
I develop a model of optimal labor income tax rates in which PUT arise endogenously from firm–worker interactions, and a corporate tax is also in place. The model shows that the corporate tax modifies the standard formula for the optimal labor income tax rate through two channels: (i) a fiscal externality, as PUT cannot be deducted from firms’ taxable profits and thus raise corporate tax revenue; and (ii) a behavioral response, as higher corporate tax rates reduce firms’ incentives to pay under the table and narrow the scope for workers to evade, thereby lowering the elasticity of taxable income (ETI). Together, these effects imply that the optimal labor income tax rate increases with the corporate tax rate. Empirically, I compare administrative payroll records with household surveys in Peru. Because the two sources cannot be linked, I use optimal transport to compute the smallest amount of PUT consistent with both wage distributions. I find that at least 13.2% of private formal employees receive PUT, and those who do are paid at least 9.7% of their reported wage under the table. Exploiting two reforms to Peru’s Agricultural Regime, I show that the ETI is large when only the labor income tax rate rises and much smaller when the labor income and corporate tax rates rise together, consistent with the model’s predictions.

This paper studies the optimal payroll tax when social insurance is transferable across spouses, so that one formal worker’s contributions also cover a partner with no contributory coverage of their own. Transferability changes the standard formula for a tax on participation in two ways. The first is efficiency: it changes what a worker who leaves formality costs the budget. When a provider, whose partner has no coverage of their own, leaves, the budget loses less than the standard formula assumes, because the whole household loses coverage. When a second spouse leaves, it loses more, because they keep coverage through a formal partner. The optimal tax is therefore higher than the standard formula says when most workers at the margin are providers, and lower when most are second spouses. The second is equity. With transferability, a household with one formal worker gets two people covered for one contribution, and these households are often poorer. This raises the optimal tax too.
Using a payroll tax subsidy for low-income women in Chile, and comparing eligible women married to formal and informal men, I estimate that about two thirds of the women at the margin are providers. In a calibration with full transferability, as in Chile, transferability raises the optimal payroll contribution of low-income couples by 3.5 percentage points. Most of this is lower efficiency cost: 3.0 points come from the fact that most of their marginal workers are providers.

Recent incidence analyses find the VAT progressive because low-income people buy more informally, assuming no pass-through to informal prices. We show that pass-through to an informal price is the formal input share of that price times the pass-through of its formal input. Peru's 2022 chicken exemption fell on informal vendors' formal input in some regions only, because supply chains differ. Where it did, informal prices fell about as predicted; elsewhere, they showed no sustained response. Assuming no pass-through to informal prices would make the VAT on purchased food steeply progressive relative to expenditure; with our model, it is mildly regressive.

This study argues that current heterogeneity in female labour force participation across elevation levels in Peru may be explained by differences in historical Spanish influence. Before conquest, men and women were equally-important in the economic sphere. After conquest, women became ”legal minors” without enough reasoning power to make economic decisions. Because of Spanish reluctance to settle at higher elevations, I argue that inclusive pre-conquest gender roles persist in those areas. Using elevation as an instrument for Spanish influence, I find that living in a district less exposed to Spanish influence significantly increases fe- male labour force participation. Results are highly robust to the inclusion of a variety of demographic, geographical, contemporary and historical covariates.

The Inca Empire was the last of a long series of highly developed cultures in pre-colonial South America. It stretched across parts of the current territories of Argentina, Bolivia, Chile, Colombia, Ecuador and the whole of Peru. The Inca Road was its 30,000-kilometer-long transportation system. The aim of this study is to identify its long-term impact on current development in Peru. Our results show that the long-run effect of the Inca Road includes increases in wages and educational attainment, a reduction of child malnutrition and an increase in children’s mathematics test scores. We also find that these effects are around 20% greater for women and explore the mechanisms that may account for this pattern.