Michael Walker


Job Market Paper

Informal Taxation and Cash Transfers: Experimental Evidence from Kenya

Abstract: Informal taxation, whereby households contribute to public goods outside the formal tax system, plays an important role in local public good financing in many low-income countries, yet little is known about its magnitude or incidence. Informal taxation is implemented by local leaders and trades off information advantages with potential elite capture and reduced enforcement. In contrast to formal tax systems, it is unclear how household informal tax payments respond to changes in income. This paper uses panel data on households and local leaders, combined with exogenous variation in household income from a large-scale randomized controlled trial of a one-time unconditional cash transfer to poor households, to study how informal taxation and public goods provision responds to household income shocks. The (temporary) cash transfers are not captured by local leaders: I find no effect on household informal tax payments, and recipient household payments are in line with their pre-treatment income. Informal taxes do respond to non-experimental changes in permanent income in panel data. Recipient households pay more formal self-employment taxes, though the magnitude of the increase is small relative to the transfer amount: less than 1 percent of total transfer income is captured by formal or informal taxes. I find no effects of the cash transfers on public goods provision. This suggests local leaders emphasize equity considerations by exempting these cash transfers to poor households from taxation, but miss out on opportunity to meaningfully increase public goods investment.

Work in Progress

General Equilibrium Effects of Cash Transfers in Kenya (with Johannes Haushofer, Edward Miguel and Paul Niehaus).

Abstract: Redistribution via cash transfers is growing in popularity as a tool for poverty alleviation due to their lower administrative costs and the increased flexibility they offer beneficiaries. Recent studies have found positive effects for individual households, but little is known about the spillover effects of redistribution in cash and its general equilibrium effects on the local economy. We study general equilibrium effects on both poor and less-poor households, enterprises and market prices using a two-level randomized controlled trial of a large-scale NGO unconditional cash transfer program to poor households in rural Kanya. Treatment is assigned at the village level, and treatment intensity is varied experimentally over clusters of villages. We utilize these helicopter drops of cash to provide insights into the macro, trade, and urban economics literatures.
Status: Data collected AEA Study Registry

How Cash Transfers Shape Citizen-State Interactions (with Kate Orkin)

Abstract: How does a short-term spike in income from a cash transfer from a foreign NGO affect interactions between local politicians and households, both recipient and non-recipient? We have randomly allocated the roll-out of GiveDirectly’s unconditional cash transfer programme in two trials across 1,069 villages in Western Kenya. We test whether, as our qualitative work suggests, recipient households that receive a positive income shock make fewer requests for financial assistance from local politicians. We also explore the governance implications by investigating households’ political attitudes and changes in participation of communities in consultations about local public goods. Finally, we examine the effect of cash transfers on the behaviour of politicians: we expect them to only claim limited credit for the transfers, and will test whether they change their amount of private assistance to recipient and non-recipient households.
Status: Survey in field


Repo runs: evidence from the tri-party repo market (with Antoine Martin and Adam Copeland), 2014. Journal of Finance 69(6): pp.2343-2380.

Abstract: The repo market has been viewed as a potential source of financial instability since the 2007 to 2009 financial crisis, based in part on findings that margins increased sharply in a segment of this market. This paper provides evidence suggesting that there was no system-wide run on repo. Using confidential data on tri-party repo, a major segment of this market, we show that, the level of margins and the amount of funding were surprisingly stable for most borrowers during the crisis. However, we also document a sharp decline in the tri-party repo funding of Lehman in September 2008.