Yifei Liu

Research

Working Papers

  • Permitting, Litigation Risk, and Energy Infrastructure Investment (Job Market Paper)
    Abstract Capital investment in energy infrastructure is essential to support economic growth and meet rising electricity demand. In the U.S., however, legal risks arising from the infrastructure permitting process may deter such investment, though their magnitude and mechanisms remain unclear. Using novel litigation data on environmental and land-use permits, I study this question in the context of renewable energy infrastructure. I find that litigation influences market entry through two channels. Directly, historical litigation deters renewable market entry by 4 percent at the mean entry rate through perceived risk, while legal precedent encourages entry by 9 percent by clarifying legal standards. Indirectly, through regulatory agency responses, litigation extends permit review timelines by 21 days on average and by 206 days following negative rulings, while legal precedent alleviates these delays through the same mechanism. Legal precedent functions as a public good: individual developers bear full costs of legal proceedings while benefits from legal clarity accrue to all future projects. This likely leads to underinvestment in legal precedent, consistent with economic theory and observed patterns where most developers with unfavorable rulings do not engage in further legal proceedings. I develop a structural model of entry and legal proceedings that accounts for selection and estimates average permitting costs of $5.5 million, or 14 percent of expected project net profits. Counterfactual simulations show that a legal fee shifting scheme would increase market entry by 6.1 percent, compared to 3.4 percent from permitting cost reductions. Internalizing the externalities of legal precedent may accelerate renewable deployment more effectively than administrative reforms alone.
  • An Empirical Analysis of the Interconnection Queue (with Sarah Johnston and Chenyu Yang) [NBER Working Paper] Reject and Resubmit at Econometrica
    Abstract Generators applying to connect to the U.S. power grid go through an interconnection queue. Most wind and solar generators that begin the process do not complete it. Using new data, we find that a long queue increases the average waiting time, and high interconnection costs are a key factor in a generator’s decision to withdraw. We develop and estimate a dynamic model of the queue and quantify the effects of policy reforms. Our simulations indicate that reducing waiting times can significantly increase completions. An alternative queuing mechanism can therefore increase completed capacity by removing certain generators to reduce congestion. A flat entry fee has a similar effect. We also quantify the effects of reforming how interconnection costs are assessed. These policy reforms lead to a substantial reduction in carbon emissions.
  • Grid Connection Costs as a Barrier to Building New Generation: Evidence and Implications for Transmission Policy (with Sarah Johnston and Chenyu Yang) [RFF Working Paper] Submitted
    Abstract Meeting projected growth in electricity demand and climate goals will require building new electricity generators. These generators must connect to an increasingly congested electric grid. We collect new data on grid connection costs for the largest regional grid operator in the United States. We find that the network upgrade costs for grid connection are increasing over time across fuel types and locations. We also find that planned generators with high network upgrade costs are much more likely to be canceled. Finally, recent transmission spending by the grid operator is associated with lower network upgrade costs for connecting generators. These findings emphasize the critical role of transmission capacity in expanding electricity generation capacity.
  • Market Structure and Transmission Investments in U.S. Electricity Markets
    Abstract Meeting the growing demand for electricity and supporting the decarbonization of the power sector requires substantial investment in transmission infrastructure. This paper evaluates the impact of a specific market design, known as market dispatch, on transmission infrastructure investment. Under traditional regulation system, a single, regulated utility company supplied electricity within a region. Beginning in the late 1990s, however, many U.S. regions transitioned to market dispatch systems, where multiple utilities generate electricity and auction-based mechanisms determine dispatch quantities. Despite this shift, transmission operations remained subject to cost-of-service regulation. I exploit the staggered adoption of market dispatch across regions using a dynamic difference-in-differences design. The results show that adopting the market dispatch system leads to an average increase of $36 million in transmission investment by utilities—nearly 50% of mean investment levels. However, I find no robust evidence of increased investment on high-voltage transmission infrastructure, which is critical for integrating renewable energy and reducing greenhouse gas emissions.

Selected Work in Progress

  • Cluster Designs and Strategic Delay in Interconnection Queues (with Sarah Johsnton and Chenyu Yang)
    Abstract Cluster study designs are increasingly used to manage grid connection requests submitted by new electricity generators. Under this scheme, grid operators group projects together to streamline the interconnection process. We collect new data on this process for SPP, a grid operator that uses cluster studies. We find that this design leads to significant interdependence in connection costs across generators. Moreover, the estimated connection cost for a given generator tends to fall as other generators leave the queue, leading to a war of attrition. We develop a model of a generator’s decision to wait or withdraw and consider the effects of alternative policies.
  • Production Network, Wind Penetration and Environmental Impact: The Case of Texas Wholesale Electricity Makret (with Sheldon Du and Qinan Lu)
    Abstract To achieve the ambitious goal of carbon neutrality, the US government is accelerating the adoption of renewable energy and spends billions subsidizing renewable energy investments every year. However, the majority of these subsidies are production-based and do not consider the locations and interconnections of newly entering renewable generators and existing fossil fuel generators within the electricity grid. Consequently, renewable generators tend to cluster in areas where resources are rich and government incentives are the highest. This clustering could lead to suboptimal transmission congestion and generation curtailment. Productions from renewable generators at different locations may substitute productions from different sets of fossil generators. In this project, we investigate the benefits of emissions reduction from renewable generations after accounting for locational and interconnection effects as well as the changing network structure.

Pre-Doctoral Publication

  • Identifying high-priority impact areas for electricity service to farmlands in Uganda through geospatial mapping (with Rebekah Shirley, Josephine Kakande, and Mark Kagarura) Journal of Agriculture and Food Research, 2021
    Abstract This article explores the food-energy nexus in sub–Saharan Africa by studying opportunities for improved agricultural productivity through electricity access. The study fills an acknowledged data gap by using geospatial analysis to identify priority areas where least-cost electricity delivery models intersect with agricultural needs. These findings are validated by community surveys in a key farming district. The analysis finds significant areas of underserved staple and cash crop farmlands can be served through grid and mini-grid electricity access within the next ten years.