UC Davis

Information and search frictions are often cited as sources of market distortions and failures in developing economies. Information communications technology (ICT) and online platforms may help solve these problems, as evidenced by increased adoptions of mobile applications, e-commerce, and gig-economy platforms. Technology-enabled tools and marketplaces may also help address challenges faced by regulatory frameworks in developing economies, such as increased visibility on informal labor markets through digital traces. Furthermore, ICTs may also help lower the costs of providing public goods and services with limited resources and under weak institutions. In this dissertation, I address some of the challenges and opportunities that have arisen in the context of rapidly improving information communication technologies and the emergence of online marketplaces.

In the first chapter, I focus on an online-platform-enabled informal labor market for low-skill workers and address the efficacy of regulatory interventions to improve worker welfare. My co-author and I document the effects of a government-mandated price-floor policy intended to improve workers' earnings in the ridesharing market in Indonesia. We hypothesize that when a market-wide impact is accounted for, a price floor regulation may have muted impact on policy objectives or unintended consequences due to adjustment mechanisms and general equilibrium effects. We measure the causal impact of the policy by taking advantage of an exogenous variation in the policy's rollout.

We find that, on average, the policy increases the average trip price, as expected from a binding price floor. However, we do not find significant effects on the overall transaction volume, driver earnings, or wages. The results can be explained by a significant increase in labor supply, which reduces the number of transactions allocated per driver hour. We fail to find evidence that the price floor policy is redistributive; the excess labor supply comes from lower-earning drivers but does not lead to their increased earnings. We also find that the policy reduces driver productivity through increased excess supply and is driven by two margins: an increased share of less productive drivers in the workforce and reduced individual productivity due to the crowded supply side. In sum, we find that the price floor policy does not achieve its objectives at an efficiency cost.

In the second chapter, I focus on the mechanics of information and search frictions in online marketplaces via a randomized controlled trial (RCT). My co-authors and I document how an information intervention triggers strategic responses, spillovers, and adjustments to the market in developing economies. We identify a context in which there is limited publicly available price information in the used car market in Pakistan and collaborate with a major online listing platform, PakWheels.com. In our randomized intervention, we provide estimates of transaction prices privately to sellers. We then measure the impact on sellers' pricing and advertising choices, and transaction outcomes. We design the experiment such that we capture both direct and spillover effects.

We find that sellers in online markets respond to an exogenously varied information environment by adjusting their pricing and advertising behaviors and that their choices generate spillovers to other sellers. The empirical estimates show that the information intervention brings directly treated sellers' listing prices closer to our price estimates and reduces advertising usage. These adjustments by treated sellers induce spillovers to sellers who are not directly treated by increasing the page views they receive from potential buyers and improving their transaction probability. The findings point to two mechanisms: 1) effects of price information are mediated by advertising tools that could countervail effects of list-pricing choices, and 2) spillovers could propagate direct effects of information intervention via adjustments by competing sellers.

In the third chapter, I address the role that ICTs can play in improving public service provision. My co-author and I focus on the context of air quality information provision in Lahore, Pakistan, where government services are intermittent and hard to access for the general public. With the availability of low-cost air quality sensors, private citizens' groups have emerged to record and disseminate real-time air quality information via social media. We test how citizens' demand for information and beliefs about air quality depend on the provider they receive information from. We conduct a field experiment in which we provide day-ahead air pollution forecasts via SMS and make salient one of the information sources---the government vs. a private citizens group. We then track respondents' willingness to pay for the SMS service and collect a series of incentivized belief measures on service quality and preference for information sources.

We find that respondents have a high willingness to pay for the forecast service, yet not differentially so between the sources. We also find that they have a significantly higher revealed preference for the assigned source against the other. Respondents' beliefs about air pollution levels are not statistically different across treatment groups, but their belief about the government forecast error is 12% higher than for the private alternative. Our findings suggest that respondents have weak priors and malleable preferences for information sources yet expect lower service quality from the government.

In summary, I provide empirical evidence on economic agents' beliefs and choices, their spillover effects, and market forces in the context of ICTs and online platforms in developing economies. In future work, I hope to extend my insights into the following areas. First, on informal labor markets, as discussed in Chapter 1, I hope to conduct with additional causal analysis on how workers adjust to income shocks and uncertainties via RCTs. Second, on information frictions on online platforms, as discussed in Chapter 2, I hope to explore other relevant types of frictions that may limit the efficacy of online markets in developing economies, such as trust and fraud. Third, on air quality information, as discussed in Chapter 3, I hope to understand how citizens adjust their behaviors in response to air quality information, with a focus on choices in the labor market.