Customer Acquisition, Rising Market Concentration and US Productivity Dynamics (Job Market Paper) [ PDF ]
Abstract: The cost of marketing and advertising has declined enormously due to the advance of digital technologies. This paper studies the macroeconomic consequences of lower marketing cost, and finds that it is a critical driving force of several striking macroeconomic trends, including rising market concentration and productivity growth slowdown since the 1990s. I develop an endogenous growth model with product market search frictions. Firms invest in innovation and marketing to build customer base, which is a long-term asset. Then I exogenously feed in the observed large drop of marketing cost into the quantitative model and find that it accounts for 83% of the rise in market concentration, measured by the largest firm’s market share. Cheaper marketing generates a positive level effect and a negative growth effect on productivity. These two effects together explain around 1/3 of the decline in productivity growth rate and successfully captures its “first-rise-then-fall” pattern over time. Finally, I conduct a welfare analysis and find that firms tend to over-invest in marketing compared to the socially optimal allocation, which implies that welfare can be improved by a marketing tax.
Presentations: 2022 EWMES (Scheduled), 2022 Trans-Atlantic Doctoral Conference (TADC), 2022 AFA Ph.D. poster session, 2021 Asian Econometric Society Meeting, 2021 China Meeting of the Econometric Society, UCLA proseminar
Revisiting Capital-Skill Complementarity, Inequality, and Labor Share, with Lee E. Ohanian and Musa Orak [ PDF ]
NBER Working Paper (No.w28537), forthcoming at Review of Economic Dynamics, 2023
Abstract: This paper analyzes the quantitative contribution of capital-skill complementarity in accounting for rising wage inequality, as in Krusell, Ohanian, Rios-Rull, and Violante (KORV, 2000). We study how well the KORV framework accounts for more recent data, including the large changes in labor’s share of income that occurred after the KORV estimation period ended. We also study how using information and communications technology (ICT) capital as the complementary capital stock affects the model’s implications for inequality and overall model fit. We find significant evidence for continued capital-skill complementarity across all model permutations we analyze. Despite nearly 30 years of additional data, we find very little change to the original KORV estimated substitution elasticity estimates when the total stock of capital equipment is used as the complementary capital stock. We find much more capital-skill complementarity when ICT capital is used. The KORV framework continues to closely account for rising wage inequality through 2019, though it misses the three-percentage point decline in labor’s share of income that has occurred since 2000.
Presentations: Philadelphia Fed, "Labor, Firms, and Macro" Reading Group
Monopoly or Efficiency? Aggregate Impact of Mergers and Acquisitions on Macroeconomic Dynamics [ PDF ]
Abstract: "Big tech" companies dominate many industries and are very profitable, but they did not get there alone. In this paper, I develop a Schumpeterian growth model to study the macroeconomic implications of mergers and acquisitions, which is a key way for many incumbents to expand and gain profits. There are two types of acquisitions, defensive M&A in which incumbents buy out the rivals to eliminate competition, and expansionary M&A where the acquirer complements startups to do radical innovations to take over businesses from other firms. The relative gains of target firms in these two types of M&A influence startups' incentives to choose incremental or radical innovations, and thus affect aggregate productivity and social welfare. I use the model to show the effect on business dynamism of introducing a technological change that enables more firms to generate positive synergies in M&A. The surge of expansionary M&A by incumbent firms increases productivity at the beginning. However, as more and more industries are taken over by these firms, their high profits enable them to pay high prices in defensive M&A to startups doing incremental innovation. This explains the rise of M&A, the increase of average markup and concentration, and the slowdown of innovation and productivity.
Presentations: 2021 WEAI Conference, UCLA Proseminar
Working in Progress
The Transmission of Shocks Across Industries: Evidence from a Billion News Articles, with Bruno Pellegrino
Abstract: We leverage an extremely large digital database of news articles, containing 1.5 billion pieces of news from over 32,000 news sources, that have been tagged by topic and industry using artificial intelligence, to construct industry-level measure of firms' exposure to a variety of economic shocks. After showing how our measurement framework can be applied to study a multiplicity of shocks, ranging from the introduction of artificial intelligence to epidemics, we focus on a specific application. We use our database to study the causality of policy uncertainty on firm equity volatility and capital investment. We ask whether companies that operate in industries more exposed to regulations and policy shocks experience higher stock price volatility and whether they scale back capital investment in response to higher policy uncertainty. While existing data sources only allow to capture time-series variation in policy uncertainty, our data and methodology enable us to investigate the transmission of policy uncertainty shocks to firms on a cross-sectional basis; this can drastically improve our ability to identify the causal impact of policy uncertainty shocks.
Market Liquidity and Bond Issuance: Effects of the Fed’s Interventions During the COVID-19 Crisis, with Huifeng Chang
Abstract: According to the conventional wisdom, firms tend to choose shorter maturity when issuing bonds. Using microdata on corporate bond issuance during the Covid-19 crisis, however, we provide evidence that the average maturity issued is longer. We develop a model of debt maturity choice of firms in the presence of rollover risk and search frictions in the secondary bond market. We use the model to rationalize our empirical findings on bond issuance. The calibrated model can be used as a laboratory to run the counterfactual for policy analysis. We show that the Fed intervention policies improve efficiency by reducing the fixed issuance cost.