Home Antitrust and Competition The New Challenges of Assessing Big Tech’s Impact

The New Challenges of Assessing Big Tech’s Impact

0
Illustration by StudioM1, via Getty Images

Big Tech firms are facing the biggest wave of antitrust legislation in their history. Academic literature reveals the complexity of possible consequences of the new antitrust bills that are currently debated. 


A recently unveiled bipartisan package in Congress proposes antitrust reforms that could have a major impact on Big Tech companies. The package includes five bills targeting companies like Amazon, Apple, Google, and Facebook that operate digital platforms and have  a market cap above $600 billion. 

These proposals are not surprising, given the recent hearings and the voices across the aisle calling to curb market power of the Big Tech companies. Big Tech firms have faced antitrust probes in Europe and China as well. While consensus builds up that some regulation is needed, it is far less clear what the optimal regulation should be. Should Big Tech firms be broken down into smaller entities? If not, what limitations should be put on their further growth, either organic or via mergers and acquisitions? 

One of the five bills, the Platform Competition and Opportunity Act of 2021, aims to shift the burden of proof in merger and acquisition cases from the government onto Big Tech firms. According to this bill, Big Tech firms would need to prove that their new acquisitions do not reduce competition. The bill clearly makes it harder for Big Tech firms to acquire startups. The National Venture Capital Association (NVCA), an industry group representing venture capital firms, opposes this bill and believes that Big Tech firms’ acquisitions encourage entrepreneurs to start new firms. The NVCA wrote: “Barring acquisitions of companies by select acquirers would close the door to this opportunity for many startups, depressing the economics of startup investment and therefore creating a significant disincentive for new company formation, job creation, and innovation in our country.” Whether Big Tech firms encourage entry of new startups and innovations is an important question, because startups are the engine of economic growth.

The academic literature demonstrates that the impact of Big Techs’ acquisitions is not straightforward. A 2019 paper by Zhao Jin provides causal evidence that if startups are good complements to Big Tech firms’ businesses, then the prospect of getting acquired can induce entrepreneurs to launch new startups. 

To show that, the study uses the spatial dimension of startup formation. It first documents that Amazon is more likely to acquire startups or invest in startups that are located in Seattle (where Amazon’s first headquarters is located). Then, it uses two announcements of Amazon’s search for second headquarters (HQ2) to identify the effect of Big Tech firms on startup formation. In January 2018, Amazon picked 19 cities as the finalists for HQ2. Then, in November 2018, the Crystal City (VA) and the New York City (NY) were announced as the winners (the latter was canceled in February 2019). 

The study shows that entrepreneurs are more likely to establish a startup in one of the 19 cities after the first announcement, and more likely to locate in the finalist cities after the second announcement. There is also a reversal in startup entry to the 17 cities that were not picked as the winners. Both the entry and the reversal patterns exist only for Amazon-type startups, startups that are similar to those that were acquired by Amazon between 2008 and 2017. Amazon HQ2 announcements did not have a significant effect on entrepreneurs’ decision to establish non Amazon-type startups in one of the 19 cities. Thus, the tendency of Amazon-type startups to co-locate with Amazon implies that those startups try to maximize the chance of getting acquired or invested by Amazon. 

The paper goes further to identify whether a startup is likely to compete with Amazon based on the textual analysis of Amazon’s 10K description of the competitors. This classification shows that Amazon-type startups that are not competing with Amazon are more likely to be formed in Amazon HQ2 cities than Amazon-type startups that are more likely to develop competing products (e.g., have e-commerce in their keywords). Based on the evidence in this paper, the Platform Competition and Opportunity Act of 2021 will reduce entry by startups.

On the other hand, a 2021 paper by Sai Krishna Kamepalli, Raghuram Rahan, and Luigi Zingales provides a theoretical model arguing that Big Tech firms can deter new startup formation. If a new startup wants to develop a competing platform to that of a Big Tech company, it would have a smaller chance of succeeding given the network externalities and the economy of scale in the digital platforms business. The network externality increases with the number of users, this tendency of users to stay with the incumbent’s platform will reduce the stand-alone value of entering platform. 

Kamepalli et al. show the evidence that relative to the mean in the entire software sector, VC investments in startups that operate in the same space as the firms acquired by Google and Facebook drop by 40 percent, and the number of deals decreases by 43 percent in the three years following a major acquisition. Similarly, the financing of new startups in the same space decreased by 51 percent relative to the financing of all new start-ups in the software industry. This evidence suggests that a desire to be Big Tech firms’ first acquisition in a business space that complements their existing business can incentivize entry by startups, as suggested by the 2019 paper by Zhao Jin mentioned earlier. Consequent entry of more startups to the same space will be discouraged because their products are substitutes to Big Tech’s existing business after the first acquisition. In other words, the first acquisition creates a “kill zone” and disincentivizes further entry. 

These two studies highlight technology overlap (complementarity versus substitutability of products) as the main issue behind the debate on whether Big Tech firms’ acquisitions deter entrepreneurial entry. Policymakers and regulators should take technology overlap into consideration when making M&A related antitrust law. UK Competition and Markets Authority conducted an analysis covering acquisition deals involving UK startups between 2008 and 2018. The analysis looked into 160 acquisitions made by Google, 71 by Facebook, and 60 by Amazon, and found that targeted startups were from various sectors, and found that their products and services are often complementary to those of acquirers.

“The academic literature demonstrates that the impact of Big Techs’ acquisitions is not straightforward.”

Besides the entry concern, another major concern about Big Tech’s acquisitions of startup firms is that these acquisitions are done in order to terminate a competitor rather than to adopt and develop the newly acquired technology. In a recent paper, Colleen Cunningham, Florian Ederer, and Song Ma show evidence for such behavior in the pharmaceutical industry. When an acquirer’s market power is large and its product portfolio overlaps with the target’s, the acquired drug projects are more likely to be shut down, what the authors refer to as a “killer acquisition.” However, the Israeli antitrust authority investigated 21 acquisition deals made by Google, Amazon, Facebook, Apple, and Microsoft, and concluded that there is no evidence for even one “killer acquisition.” Putting aside whether tech giants frequently shut down acquired projects, it is difficult to prove that the intention behind “killer acquisition” is to preempt future competition. The discontinuation of a product may be necessary whenever further development is costlier than the expected benefit from launching the product. 

In addition, large companies’ equity investment in startups is generally viewed as an important complement to capital supply from independent venture capital firms. In a paper published in 2014 by Thomas Chemmanur, Elena Loutskina, and Xuan Tian, the authors show that unlike independent venture capital firms pursuing financial returns, large firms’ corporate venture capital (CVC) affiliations want to create strategic value for their parent companies through funding startups. Therefore, CVCs are more tolerant to riskier projects, what results in more funding to riskier but potentially more innovative startups. Moreover, there might be additional benefits for startups to get funded by large companies. Richmond Mathews wrote a paper in 2006 that argues that the equity investment can effectively reduce the incumbents’ incentives to enter into the startup’s market if the incumbent has not entered yet.

Further, Thomas Hellman wrote in 2002 about a theoretical possibility that entrepreneurs may worry that accepting investment from a large firm will result in losing their technology to the investor. His paper shows that a corporate investor is more likely to exploit a startup’s technology when the overlap between the startup’s and tech giant’s technology is larger. As a result, entrepreneurs whose technology complements that of the Big Tech firm are the one who are likely to accept investment from Big Tech venture capital arms. Overall, restricting Big Tech’s direct investment in startups is likely to negatively affect startups that have complementary technology and could benefit from the partnership with the Big Tech firms. 

Besides the impact that Big Tech firms have on startups via acquisitions and equity investment, large companies could have an indirect effect on startups as well. In a recent paper, we study 177 AI startups established by 363 graduates of 69 North American Universities. We find that when AI professors leave academia to work for industry, universities produce less AI startups and these startups raise less funding. Not only it shows that AI knowledge is critical for students’ success as entrepreneurs, but also that hiring professors indirectly reduces the number of startups. Out of 211 AI professors who partially or fully left academia between 2004 and 2018, 65 were hired by Google (23), Amazon (17), Microsoft (13), Facebook (8) and Apple (4). Each firm is unlikely to internalize the negative effect of its poaching behavior on startup formation. Overall, our research shows that Big Tech firms can disrupt startups, possibly unintentionally, many years before these startups are formed. It also highlights the complexity faced by the regulators because these indirect effects of Big Tech firms are difficult to predict and even harder to measure.   

As Big Tech firms penetrate virtually every aspect of people’s lives, they likely have additional direct and indirect effects on consumer welfare, entrepreneurship, innovation, education, and economic growth. Future regulations should consider both the costs and benefits of Big Tech on the economy-wide innovation and growth. The academic research shows that breaking up Big Tech firms could help to address some antitrust issues but could also reduce the formation of new startups, especially if these startups complement Big Tech’s existing products and services.

References

  • Cunningham, Colleen, Florian Ederer, and Song Ma. 2021. “Killer acquisitions.” Journal of Political Economy
  • Chemmanur, Thomas J, Elena Loutskina, and Xuan Tian. 2014. “Corporate venture capital, value creation, and innovation”. The Review of Financial Studies 
  • Gofman, Michael and Zhao Jin 2021. “Artificial Intelligence, Education, and Entrepreneurship”. Working Paper
  • Hellmann, Thomas. 2002. “A theory of strategic venture investing”. Journal of Financial Economics 
  • Israel Competition Authority. 2020. “Acquisitions of Israeli Start-ups: Ex-post Examination”
  • Jin, Zhao. 2019 “How do large companies affect entrepreneurship: evidence from Amazon’s HQ2 search”, Working Paper. 
  • Kamepalli, Sai, Raghuram Rajin, and Luigi Zingales. 2021. “Kill Zone”, Working Paper. 
  • Mathews, Richmond D. 2006. “Strategic alliances, equity stakes, and entry deterrence”. Journal of Financial Economics