« Back to Results

Electricity Markets and the Environment

Paper Session

Sunday, Jan. 6, 2019 1:00 PM - 3:00 PM

Atlanta Marriott Marquis, M103
Hosted By: Association of Environmental and Resource Economists
  • Chair: Mark Jacobsen, University of California-San Diego

Are Residential Electricity Prices Too High or Too Low? Or Both?

Severin Borenstein
University of California-Berkeley
James Bushnell
University of California-Davis


Advocates of market mechanisms for addressing greenhouse gases
and other pollutants typically argue that it is a necessary step in
pricing polluting goods at their social marginal cost (SMC). Elec-
tricity prices, however, deviate from social marginal cost for many
reasons, some of which cause prices to be too low{such as pollution
externalities{and others cause prices to be too high{such as recov-
ery of xed costs. Furthermore, because electricity is not storable,
marginal cost can
uctuate widely within even a day, while nearly
all residential retail prices are static over weeks or months. We
study the relationship between residential electricity prices and so-
cial marginal cost, both on average and over time. We nd that
while the di erence between the standard residential electricity rate
and the utility's average (over hours) social marginal cost is rela-
tively small on average in the US, there is large regional variation,
with price well above average SMC in some areas and price well
below average SMC in other areas. Furthermore, we nd that for
most utilities the largest source of di erence between price and
SMC is the failure of price to re
ect variation in SMC over time.
In a standard demand framework, total deadweight loss over a time
period is proportional to the sum of squared di erences between a
constant price and SMC, which can be decomposed into the compo-
nent due to price deviating from average SMC and the component
due to the variation in SMC. Our estimates imply that if demand
elasticity were the same in response to hourly price variation as
to changes in average price, then the sales-weighted average share
of deadweight loss attributable to the failure to adopt time-varying
pricing is 62%, with the remainder attributable to the gap between
price and average SMC. These deadweight loss shares, however,
vary dramatically across utilities and regions, and are sensitive to
demand elasticity assumptions.

Transmission Constraints, Intermittent Renewables, and Welfare

Jacob LaRiviere
Xueying Lu
University of California-San Diego


We use the roll-out of a large transmission expansion in Texas' electricity market to measure the market impacts of the transmission expansion on benefits of increased renewable capacity. We find large market benefits leading to a payback period of roughly 14 years. However, total welfare improvements from reduced congestion depend on how global non-market externalities are internalized by regional policy makers: accounting for non-market externalities reduces the payback period of this project from 14 to less than 9 years. We discuss the finding's implications for the welfare of regional decisions to build transmission capacity for the U.S. wholesale electricity market.

Ramping Up Renewable Energies: The Role of Ramping Cost and Electricity Storage

Haoyang Li
Michigan State University


A major factor limiting the growth of renewable energies (REs) is their variability or
intermittency, especially when facing high electricity storage costs. In response to the variable
electricity supply from renewable sources, fossil-fuel power generators have to experience more
start-up/shut down cycles and power ramping (i.e. changing output without shutting down),
resulting in large physical damages to the generators. Energy storage (ES) reduce such output
adjustment needs by smoothing out renewable supplies. Generators with different levels of
flexibility measured by start-up and ramping costs are affected differently by RE and ES
expansions. Existing studies that value RE and ES largely ignore fossil-fuel power generators’
flexibility limitations and produce inaccurate estimates of power generators’ production
decisions, emissions and profits. Therefore, optimal RE and ES capacity calculation based on
such estimates are also biased.
In this paper, I use hourly power plant generation and electricity price data to structurally
estimate generator specific start-up cost and ramping cost in Texas ERCOT grid under a dynamic
discrete/continuous choice framework. Generators’ reluctant output adjustments are explored to
identify such costs. Estimation results confirm that large flexibility heterogeneities exist among
Using the estimated cost parameters, I calculate bias in RE and ES valuation if power
generator flexibility is not accounted for. I then simulate profit changes of power generators with
different levels of flexibility as RE and ES penetration rises. These results help predict future
investment paths of different types of generators as renewable shares increase and storage
becomes cheaper. Finally, I predict social welfare gain of diversifying renewable sources by
adding solar energy into ERCOT to supplement wind energy. Compared to ES, solar energy also
smooths out wind energy supplies and is cheaper, but more variable. The results could inform the
optimal mix of REs and ES in ERCOT under current grid flexibility.

Fracking, Farmers, and Rural Electrification in India

Robert Fetter
Duke University
Faraz Usmani
Duke University


We exploit a natural experiment in India to investigate whether large-scale rural electrification is
necessary—and not, as is often implicitly assumed, sufficient—for household welfare and regional
economic growth. Smallholders in northwest India grow the vast majority of the world’s guar, a crop
that yields a potent thickening agent used in hydraulic-fracturing (“fracking”) fluid. In response to
the United States’ fracking boom, Indian guar prices increased by over 1,000 percent between 2006
and 2010. Using multiple identification strategies, we first evaluate the impacts of this exogenous
income shock on rural labor-market outcomes. Leveraging population-based discontinuities in
the contemporaneous rollout of India’s massive rural electrification scheme, we ultimately aim to
investigate the extent to which these impacts are confined to areas electrified before the boom.
Using synthetic controls applied to two decades of satellite-detected nighttime lights for nearly
600,000 Indian villages, we show that the US fracking boom led to sizable increases in nighttime
brightness in India’s guar belt. Nighttime luminosity is correlated with regional economic activity
(Chen and Nordhaus, 2011; Doll et al., 2006; Henderson et al., 2012), which suggests potentially
transformative changes taking place on the ground. Indeed, using multiple rounds of the Indian Census,
we demonstrate a dramatic increase in the share of farmers in the rural population accompanied
by an increase in the daily wage-rate for agricultural laborers.
To what extent are the benefits associated with this exogenous boom driven by access to grid
electricity? In ongoing analyses, we look to exploit a village-level population-based eligibility
threshold in the roll-out of India’s rural electrification scheme as part of a regression discontinuity
design. We hypothesize that guar-growing villages that were electrified prior to the boom are better
equipped to exploit the increase in economic opportunity associated with the guar boom than those
that were not. Consequently, we hope to shed light on the foundational role of electrification in
Mark Jacobsen
University of California-San Diego
Sarah Johnston
University of Wisconsin-Madison
Steven Puller
Texas A&M University
Anant Sudarshan
University of Chicago
JEL Classifications
  • Q4 - Energy
  • Q5 - Environmental Economics