UC Davis

This dissertation uses econometric methods to explore the interaction between utility policy and climate change on consumers. Energy and water utilities play an important role in all consumers' lives. As such, the decisions that regulators make with respect to these services can have large ripple effects through the economy. It is important for regulators to understand the full impact of their policy decisions. The difficulty of these decisions is further compounded by climate change. Both greater variation in temperatures and changes to historical precipitation patterns will create difficult trade-offs for policy makers. I leverage multiple large data sets in order to address these trade offs.

The first essay looks at the impact of extreme heat on consumer credit scores. A change in a consumer's credit score can have a wide ranging influence on their financial opportunities. Household energy bills are the most obvious example of how extreme temperatures can impact credit scores. While we stay agnostic about the exact causal mechanism, the results of this analysis reflect patterns important to recognize in utility policy. We take a random nation-wide sample of credit score data from 500,000 consumers over a nearly 10 year period and pair it with ZIP code level weather data to estimate the impact of unexpected temperature fluctuations on consumer credit scores. These temperature spikes are unlikely to be anticipated by consumers since they fall outside of weather norms for the quarter and ZIP code. As a secondary analysis, we use a distributed lag model to see how the effect of temperatures influences credit scores over time. We find that high temperatures lead to modest, but tangible, decreases in credit scores on average. We also find that these effects linger, but dissipate over the course of a year. The average effect of this decrease is relatively modest, however our analysis does not preclude larger heterogeneous effects. To address these outcomes, lawmakers can create policies that target the immediate short-term financial difficulties and long-term impacts to consumer credit following extreme temperature spikes.

The second essay measures the outcomes and cost-effectiveness of four Californian water demand management (WDM) programs. WDM programs aim to reduce water consumption, and by doing so, also reduce energy use and greenhouse gas (GHG) emissions. Energy is consumed by water utilities to pump, convey, distribute, and treat water and by consumers to heat water. WDM programs are widely used, but the outcomes predicted by the governments and utilities implementing them are sometimes based on optimistic assumptions about the efficacy of the upgraded appliances. Rather than relying on ex ante deemed savings approaches, we present ex post empirical estimates of the water and energy savings achieved by four urban water retrofit programs. We start by estimating the water and energy savings incurred by the programs to evaluate their cost effectiveness, then look at the distinct cost savings for households, utilities, and government agencies. We find that these programs often saved much less water and energy than was estimated by deemed savings methods. Therefore, the programs where only sometimes cost effective on a household level and not cost effective for utilities based on internal cost savings. This result largely stems from how water is priced in California. Additionally, only one of the four programs was cost effective in reducing energy usage and GHG emissions as it reduced the amount of water heated by consumers while the other programs focused on reducing cold water consumption.

The third essay summarizes various methods for evaluating water and energy savings programs. Government agencies, utilities, and community organizations have used water and energy savings programs in order to conserve resources and delay capital investments for many years. These programs are often evaluated ex ante or ex post in order to determine their effectiveness. Often, these evaluations are done internally by the same organizations which administer the programs. As discussed in the second essay, these methods can sometimes rely on strong assumptions about consumer behavior and the efficacy of upgraded fixtures. This paper explains the assumptions made by various popular evaluation methods and describes in which settings these methods can be appropriately applied and what data they require. This paper serves as an aid to program administrators so that they can better plan, execute, and evaluate water and energy saving programs.