« Back to Results

Innovation and Production

Paper Session

Sunday, Jan. 5, 2025 10:15 AM - 12:15 PM (PST)

Hilton San Francisco Union Square, Union Square 13
Hosted By: Econometric Society
  • Chair: Yueran Ma, University of Chicago

Automation-Induced Innovation Shift

Lin William Cong
,
Cornell University
Yao Lu
,
Tsinghua University
hanqing shi
,
Tsinghua University
Wu Zhu
,
Tsinghua University

Abstract

We study the impact of exposure to automation on corporate innovation, which informs how innovation begets innovation. We document that firms with high robotics exposure witness a decline in technology similarity over time and significantly shift innovative activities toward AI, which automation intuitively complements. The shift is more pronounced for firms with greater AI-related research experience or generate more data. Furthermore, AI patents are more costly than non-AI patents in various dimensions, including team size, researchers' labor input and inventors' originality. Consequently, firms with high automation experience a significant rise in R&D expenditure but a decline in the number of new patents in subsequent years. Against the backdrop of rising automation, this explains the puzzling observation in the literature that at the aggregate level, firms seem to become less innovative in recent years despite greater R&D expenditures. Finally, we present a simple dynamic equilibrium model to rationalize such innovation shifts.

Value without Employment

Simcha Barkai
,
Boston University
Stavros Panageas
,
University of California-Los Angeles

Abstract

Young firms’ contribution to aggregate employment has been underwhelming. However, a similar trend is not apparent in their contribution to aggregate sales or aggregate stock market capitalization. We study the implications of the arrival of “low marginal - high average” revenue-product-of-labor firms in a stylized model of dynamic firm heterogeneity, and show that the model can account for a large number of facts related to the decline in “business dynamism.” We study the long-term implications of the decline in business dynamism on the economy by providing analytical results that connect the decline in dynamism to the eventual decline of consumption.

Superstars or Supervillains? Large Firms in the South Korean Growth Miracle

Jaedo Choi
,
Federal Reserve Board
Andrei A. Levchenko
,
University of Michigan
Dimitrije Ruzic
,
INSEAD
Younghun Shim
,
International Monetary Fund

Abstract

We quantify the contribution of the largest firms to South Korea’s economic performance since 1970. Using firm-level historical data, we document a novel fact: firm concentration rose substantially during the growth miracle period. To understand whether the increased importance of large firms contributed positively or negatively to the South Korean growth miracle, we build a quantitative heterogeneous firm small open economy model. Our framework accommodates a variety of causes and consequences of (changes in) firm concentration: productivity, distortions, selection into exporting, and oligopolistic and oligopsonistic market power in domestic goods and labor markets. The model is implemented directly on the firm-level data and inverted to recover the drivers of changing concentration. We find that most of the increased concentration is attributable to higher productivity growth of the largest firms. Shutting down differential productivity growth of the top 3 firms within each sector would have decreased firm concentration but nonetheless would have reduced welfare by 2%. Differential distortions and foreign market access of the largest firms played a more limited role in the trends in concentration and had a smaller welfare impact. Thus, the largest Korean firms were superstars rather than supervillains.

Teaching Economics to Machines

Hui Chen
,
Massachusetts Institute of Technology and NBER
Yuhan Cheng
,
Tsinghua University
Yanchu Liu
,
Sun Yat-sen University
Ke Tang
,
Tsinghua University

Abstract

Structural models in economics often suffer from a poor fit with the data and demonstrate suboptimal forecasting performances. Machine learning models, in contrast, offer rich flexibility but are prone to overfitting and struggle to generalize beyond the confines of training data. We propose a transfer learning framework that incorporates economic restrictions from a structural model into a machine learning model. Specifically, we first construct a neural network representation of the structural model by training on the synthetic data generated by the structural model and then fine-tune the network using empirical data. When applied to option pricing, the transfer learning model significantly outperforms the structural model, a conventional deep neural network, and several alternative approaches for bringing in economic restrictions. The out-performance is more significant i) when the sample size of empirical data is small, ii) when market conditions change relative to the training data, or iii) when the degree of model misspecification is likely to be low.

Discussant(s)
Anastassia Fedyk
,
University of California-Berkeley
Lukas Kremens
,
University of Washington
Yongseok Shin
,
Washington University-St. Louis
Sai Ma
,
Federal Reserve Board
JEL Classifications
  • G3 - Corporate Finance and Governance