UC Berkeley

Financial incentives are traditionally relied upon by economists to influence behavior, however there is increasing evidence that non-market-based solutions may be preferable. This outcome particularly emerges when dealing with inattentive individuals or in situations where financial approaches face prohibitive legislation, budgetary restrictions, or political pushback. Understanding the impact of using non-market-based behavioral instruments, and how they compare to the use of market-based financial incentives, is crucial for governments to design effective public policy, as well as for private institutions to maximize profit. In this dissertation, I consider the setting of drought-prone California and its water consumers. The first two chapters contribute to the behavioral and environmental economics literature by exploring the impact of using non-market-based techniques – namely the use of information provision and moral suasion – in encouraging water conservation. The last chapter highlights the importance of finding effective approaches to water conservation by quantifying the impact of drought on economic outcomes.

In the first chapter, I use a novel natural experiment and a randomized field experiment to investigate how information and financial incentives compare in influencing the behavior of inattentive customers. Specifically, I study how these two tools compare in encouraging water customers to fix in-home leaks. I find that information is a powerful, low-cost tool in swaying behavior, and can be even more effective than financial incentives. Importantly, the impact of a financial incentive depends on a customer’s typical bill-to-bill charge variance, such that high-variance customers are less likely to react to a financial incentive and respond more to a clear informational signal. Financial incentives under one standard deviation of customer bill-to-bill variation likely go unnoticed. In the observed setting of customers with in-home leaks, the average customer has a 30% relative standard deviation of the month-to-month bill and only responds to a bill increase of 50% or more. Further, the impact of financial incentives vary by customer income and are considerably less effective on customers with automatic bill payment. I also find that while sending information by mail is effective, delivery through email or text may be preferable in time sensitive situations and for encouraging the use of online resources.

In the second chapter, I assess the impacts of using moral suasion via public appeals to encourage behavior change. As a case study, I analyze the effect of Governor Jerry Brown’s public pleas for water conservation in the face of California’s record-breaking drought. Using high frequency hourly consumption data at the household-level for the years 2012-2015, I conduct an event study to understand the level of short-term water conservation associated with these appeals. I find statistically and economically significant decreases in water consumption in the single-family residential sector of San Francisco in the two weeks following a well-publicized public appeals announcement. These short-term decreases range from 1.9 - 4.6% of total single-family residential water demand.

In the third chapter, co-authored with David Sunding and Maximilian Auffhammer, we evaluate the effect of the drought on economic outcomes. In this study we use ex post impact assessment methods to measure the effect of drought on farm employment and harvested acreage for the fifth largest economy in the world – California. We find evidence of a statistically and economically significant relationship between surface water imports and both employment and harvested area. We also present evidence that the effects of drought are smaller in areas with better access to local water supplies, especially groundwater, and have declined over time. The latter observation is consistent with observed shifts in land allocation toward perennial crops and with increased reliance on groundwater extraction, particularly in dry years. These trends may not be sustainable in light of the State’s recent efforts to curb groundwater overdraft. Our results suggest that absent other interventions, the future effects of drought on economic outcomes in California agriculture could be even larger than those observed in the recent past.