Changing Population Exposure to Pollution in China’s Special Economic Zones

Martin and Zhang (2020) show that the increase in manufacturing output in China's special economic zones significantly increased air pollution. In this complementary paper, we document that widespread migration into these industrial clusters between 2000 and 2010 did little to change overall population exposure to pollution. We show using satellite air pollution data that there is much heterogeneity across zones and that most population growth occurred in relatively cleaner zones, supporting recent literature that documents local willingness to make location decisions based on environmental quality (Chen et al. 2019, Khanna et al. 2020).

Inference for Support Vector Regression under ℓ1 Regularization

We provide large-sample distribution theory for support vector regression (SVR) with l1-norm along with error bars for the SVR regression coefficients. Although a classical Wald confidence interval obtains from our theory, its implementation inherently depends on the choice of a tuning parameter that scales the variance estimate and thus the width of the error bars. We address this shortcoming by further proposing an alternative large-sample inference method based on the inversion of a novel test statistic that displays competitive power properties and does not depend on the choice of a tuning parameter.

The Common Application and Student Choice

We study the effects of the Common Application (CA) platform, which allows students to submit a single application to multiple institutions, on student choice. Using individual-level data from freshman surveys over the period 1982-2014, we develop two proxies for student choice, one based upon the number of applications submitted and another based upon students attending non-first-choice institutions. Using these proxies, we first document sharp increases in student choice over time. Linking these outcomes to the timing of CA membership, we provide evidence of a link between CA entry and increased student choice.

Microcommitments: The Effect of Small Commitments on Student Success

We conducted a randomized control trial with nearly 1,000 students to test whether using a text-message-based tool that provides small commitments with social accountability helps students procrastinate less and maintain engagement with course material, yielding improved academic performance. All students received identical content via text message, but the treatment group engaged with an online platform encouraging commitments and featuring a social feed with rankings. The analysis reveals that microcommitments with social accountability improved academic performance relative to students who received simple nudges. The increase is equivalent to 3 percentage points on an exam, driven by online and hybrid classes.

Adaptive Correspondence Experiments

Correspondence experiments probe for discrimination by manipulating employer perceptions of applicant characteristics. We consider the gains from dynamically adapting the number and quality of fictitious applications each employer receives to their prior callback decisions. Calibrating employer behavior to experimental data from Nunley et al. (2015), we find that it is possible to cut the number of applications required to detect a fixed number of discriminators roughly in half relative to a benchmark design with a fixed number of applications per job. These gains are achieved primarily from abandoning jobs with very low callback probabilities and those that call back Black applicants.

Product Proliferation under Rational Inattention: Application to Health Insurance

In markets with complicated products such as insurance, why do firms offer many products even when consumers appear to receive little benefit? We show that when consumers face information acquisition costs, firms may have an incentive to introduce many undifferentiated products. This allows firms to gain market share and increase markups. We document initial evidence consistent with the model using data from Medicare prescription drug insurance. Insurers that offer more duplicate or similar plans have higher-cost plans. These results suggest a role for policymakers to restrict product proliferation in markets with complicated products.

Nominal Wage Adjustments during the Pandemic Recession

In this paper, we show that the pandemic recession has led to frequent cuts in nominal wages. Within three months in 2020, as many wage cuts had occurred as occurred throughout the Great Recession. Unlike employment declines, wage cuts were concentrated at the top of the wage distribution. However, these cuts have been relatively short lived, particularly among high earners. Finally, wage cuts have been concentrated in firms that have seen large employment declines. Wage cuts appear not to be a substitute for cutting employment, at least when the shock to labor demand is this large.

Could Gentrification Stop the Poor from Benefiting from Urban Improvements?

When policymakers invest in urban infrastructure, there are concerns that poor residents living near the infrastructure will be displaced. This paper investigates mechanisms that may lead to such infrastructure-induced gentrification using a general equilibrium urban commuting model. Our goal is to elucidate the channels through which infrastructure-induced gentrification occurs and understand how policy choices mitigate or accentuate gentrification.

Maintaining and Enhancing Labor Statistics during the COVID-19 Pandemic

The US Bureau of Labor Statistics (BLS) continued to produce key employment statistics while reengineering operations as a result of the COVID-19 pandemic. In-person data collection moved to electronic, deadlines were adjusted, and flexibilities were implemented to get the job done, all while maintaining confidentiality and data quality. Opportunities arose, such as collection of new information from households and businesses on changes that resulted from the pandemic. This is a story that continues to unfold and will transform the way BLS moves forward. This paper focuses on changes in BLS operations and in the economy as reflected in BLS data.

Measuring Consumer Spending Using Card Transaction Data: Lessons from the COVID-19 Pandemic

Pandemic-driven economic changes are measurable in days and weeks rather than months and years, underscoring a need for more timely economic data to inform the public. We summarize newly available estimates of spending using card transaction data that are available on the Bureau of Economic Analysis website and analyze correlations between card data and official series. We find that card data perform well in measuring large changes in the economy around the pandemic, although the correlations are much lower during a period of stability prior to the pandemic. This pattern is likely attributable to a shift in the signal-to-noise ratio over these periods.