Print
E-MailWei Jiang is Arthur F. Burns Professor of Free and Competitive Enterprise in the Finance Division at Columbia Business School; Yuehua Tang is Emerson-Merrill Lynch Associate Professor of Finance at University of Florida Warrington College of Business; and Vincent Yao is AREA Professor of Real Estate at Georgia State University J. Mack Robinson College of Business. This post is based on a recent paper by Ms. Jiang, Mr. Tang, Mr. Yao, and Rachel Xiao.
Advances in financial technology (fintech) are reshaping the landscape of financial services in the United States and globally. The term “fintech” refers to technology and innovation that aim to compete with traditional methods and channels for the delivery of financial services. Telecommunications and information technology have been adopted by financial service providers to create new options and to ease access by consumers (households and businesses) to navigate the complexity and constraints they face. Although the term has gained its prominence in the recent decade as an external disruptor, we are reminded that the evolution of finance has always worked in tandem with the adoption of new technologies, from wire transfer as a long-distance payment technology in the late 1800s to credit cards and automated teller machines (ATMs) during the 1950s and 1960s. Post-financial crisis has marked a dramatic shift toward decentralization (e.g., blockchains and crypto-asset) and disintermediation (e.g., peer-to-peer lending platforms), imposing disruption on the established financial institutions.
Economists have also long debated the trade-off between the new opportunities for businesses and consumers from technological advancement and the labor force displacements caused by them. The common empirical challenge to quantify the effect of technologies on jobs and firm’s outcomes is due to the general lack of ex ante measures for exposure to technology at the micro-level. Our study focuses on such relationship in the context of fintech innovations, and our first objective is to overcome the challenge by developing a novel measure of occupation exposure to fintech innovations. Such a measure is constructed by cross-analyzing and extracting the similarity in the textual information in job task descriptions and that in recent fintech patent filings. Specifically, our fintech exposure measure captures both the similarity between the two text corpuses (i.e., job task descriptions and fintech patent filings) and the intensity of fintech innovations (e.g., the amount of fintech patent filings). The procedure results in time-varying fintech exposure scores for the universe of 772 occupations as classified by the six-digit O*NET Standard Occupation Code (SOC), which can also be aggregated to the firm or industry level.
The second, and main objective of our study is to characterize and quantify how demand for talent shifts in response to fintech shocks. To this end, we link job postings by firms (and the states they reside in) from Burning Glass Technology (BGT) to individual occupations, and then to our measure of exposure to recent fintech innovations. The resulting panel consists of about 300,000 cohorts at occupation-state-year level, aggregated from the original 161.6 million BGT-listed vacancies during 2007, and 2010-2018. We find that the job posting of occupations in the top quartile of fintech exposure (“the most exposed” hereafter) experienced significant drops (as a share of all job postings in a given state and year) during the sample period. After controlling for state by year fixed effects and competing technology exposures (from AI and software), we find that the most exposed occupations experienced a 5 percent loss of job posting shares from 2007 to 2018, confirming a disruptive effect of the technology on jobs. Among all subfields of fintech innovations, data analysis, blockchain, and robo-advising have the greatest effects.
The loss of jobs exposed to fintech is not evenly borne across industries, firms, and geography. Three industries most exposed to fintech innovations, including finance, professional, management and administrative services (PMA), and information, accounted for 40% of all job postings in the U.S. in 2007 but have lost nearly 13 percentage points by 2018. Likewise, traditional financial hubs, such as New York metro, Boston (MA), Washington metro (DC, MD and VA), Charlotte (NC), Atlanta (GA), Chicago (IL), San Francisco (CA), Seattle (WA) and state of Texas have suffered the steepest losses of jobs that are most exposed to fintech. Also, we find that fintech innovations are concentrated in four industries: finance, information, manufacturing, and PMA. This pattern is confirmed by further finding that the financial industry is both the target of disruption and a leader in the fintech innovation effort, and that financial firms are both inventing and acquiring fintech patents more than others. This is contrary to a conventional belief that fintech innovation is primarily sourced outside the finance and related industries.
Firms are not expected to be passive players in a wave of disruption. We examine one aspect of their response, namely, the change in their recruiting strategies for jobs that have been overall downsized in relation to fintech exposure. Firms resort to upskilling in hiring of fintech-disrupted jobs, requiring more education attainments and longer work experiences. The demand for “finance + software” skills and “software-only” skills rises as fintech exposure increases, but that for “finance-only” skills goes in the reverse direction. However, firms’ human resource adjustments are limited by local market conditions. Ample supply of quality labor and light labor protection regulations help firms weather the disruption better. Based on the Herfindahl-Hirschman Index (HHI) constructed at occupation level, we also find that jobs exposed to fintech become more concentrated across industries and states, suggesting that workers associated the peripheral players (in terms of both industries and regions) are the most vulnerable to the technology shock.
A disruptive force on jobs due to technology does not speak to its impact on the operating outcomes of firms, e.g., in terms of sales growth and returns to capital. Therefore, the last main objective of the paper is to shed light on how firms fare when facing fintech exposure. Though the most exposed firms indeed experience significantly lower employment growth relative to other firms, confirming the relation at the occupation level, they do not suffer in sales growth and return on assets (ROA), nor in research and development (R&D) investment. In fact, inventor firms (i.e., firms that are the original developers of the fintech patents), but not acquisition-driven innovating firms (i.e., firms that acquire fintech patents), are the bright spots on the landscape: they hire more, invest more in R&D, and enjoy higher sales growth and return on assets. In sum, fintech constitutes a disruptive force for workers but not for (shareholders of) firms, and there is a win-win situation at firms that are originators of new technology.
In summary, this study aims to inform the ongoing debate in whether and how fintech constitutes disruptions and/or presents the growth opportunities, especially with respect to labor demand and employment. We discover that job postings in the most exposed occupations suffer a significant decline both in absolute magnitude and relative to other occupations. The exposed firms resort to upskilling (in terms of the requirement of skills, experience and educational attainments) in hiring albeit among overall downsizing. Fintech-exposed jobs also become more concentrated across industries and states. Nevertheless, innovative firms and finance sector manage to offset the economy-wide negative impact to different degrees. Finally, firms producing original fintech innovations themselves gain in both employment and operating performance.
The complete paper is available for download here.
