Chairwoman Velázquez, Ranking Member Chabot, and members of the Committee. Thank you for inviting me today to talk about large technology companies’ impact on small business. I am Joe Kennedy, Senior Fellow with the Information Technology and Innovation Foundation (ITIF). ITIF is the world’s leading think tank for science and technology policy and focuses on developing better policy to promote innovation, increase competitiveness, and improve productivity. In my capacity at ITIF I have written on Internet platforms, antitrust, the gig economy, and other issues.
The small businessperson has long played an iconic role in American society, symbolizing independence, fortitude, and diligence. However, both the challenges and opportunities facing small businesses have grown in an increasingly global economy with a ubiquitous Internet that lets consumers quickly compare prices, find reviews, and shop online. On the whole, Internet platforms play a positive role in helping businesses of all sizes deal with these changes. More importantly, by delivering efficiencies and increasing competition, they benefit consumers by lowering prices, leaving them with more money to spend on other goods and services.
In cases where large technology platforms compete fairly with small businesses and gain market share, this is something that policymakers should welcome, not oppose, for by definition such results boost consumer welfare and economic growth. However, when the conduct of large technology companies, particularly technology platforms, raises legitimate antitrust issues, these problems can be adequately addressed by applying existing laws. We should not enact new, restrictive laws or engage in new antitrust practices simply to rein in some of America’s most productive companies. Nor should we try the Herculean task of trying to break them up. Both would come with the consequence of less innovation, productivity, and consumer welfare.
Most large technology companies, including Internet platforms, give business owners tools they need to succeed. The most challenging aspect of any new business involves finding customers, establishing a reputation, and setting up the fixed costs involved in running a business. These latter include accounting, computer systems, human resources, and complying with the many federal, state, and local regulations. Platforms help small companies with all of this in the following ways:
- Internet platforms allow any business to set up a website, Facebook page, Instagram account, YouTube channel and other online presences to reach customers, exchange information, and build a loyal customer base. Prior to these platforms, most small businesses were limited to local media, which was geographically limited, comparatively expensive, and did not provide any feedback about effectiveness.
- Firms such as PayPal and Square make it easier to handle payments, greatly reducing the risk of nonpayment and fraud, and integrate directly with various accounting software systems, reducing administrative overhead.
- Platforms such as TaskRabbit and Thumbtack match businesses with customers, saving them some of the time and effort involved in advertising. They also help small companies find workers to deal with sudden increases in volume.
- Rating systems on sites such as eBay and Yelp help companies build reputations for excellence and provide transparency in markets.
- Cloud services from Amazon, Dropbox, Microsoft and other companies provide easily scalable computer capacity at low cost. These services also offer best-in-class security features, allowing small businesses to protect consumers’ information from hackers and reduce downtime and lost business from technical problems.
- Internet platforms allow even the smallest company to targeted advertising to those most likely to be interested in its products, which makes effective marketing more affordable for small companies.
- Finally, marketplaces, such as Amazon, eBay, and Etsy, give even the smallest companies the ability to sell online to consumers across the world. They also handle logistics and delivery, saving the company from having to perform these services itself and allowing it to concentrate on customer service and the quality of its products.
Online platforms therefore significantly reduce the costs of market entry for businesses and enable entirely new business models. This success depends on their flexibility to design rules and terms of service that maximize their value to consumers and producers.
Many of the most important battles in the technology industry do not involve small business. Whatever the merits of Spotify’s antitrust suit against Apple, the dispute involves two large and highly successful firms that can look out for themselves. But it is also important to acknowledge that some Internet companies compete with small business. This is neither new nor disturbing. With the growth in national and now global markets, the optimal size of businesses in many industries has grown. As ITIF President Robert Atkinson and Michael Lind point out, with growing markets, large companies are often more efficient than their competitors. In competitive markets, the lion’s share of the benefit from efficiency is passed on to consumers. The effect of enhanced competition can reduce the market share of smaller, less efficient rivals. Where this displacement is the effect of legitimate competition, public policy should encourage it, or at minimum not hinder it. The purpose of antitrust policy should be to enhance consumer welfare and ensure that businesses do not engage in anticompetitive conduct. It should not be to protect companies from the legitimate effects of competition.
One aspect of competition occurs when companies that operate a platform also offer their own products over that platform. The fear is that the platform owner will compete with the most profitable products offered by others. Some commentators have expressed concern that this would allow for unfair competition. For example, Senator Elizabeth Warren (D-MA) has proposed prohibiting platform owners from competing with companies that use their platform.
But once again, these practices are neither strange nor new. Walmart, Safeway, and CVS, for example, all offer shelf space to thousands of brands. In many cases they use in-store analytics to identify what products they should sell under their own brand, competing with the brands they already sell. For example, Costco’s Kirkland brand directly competes with those of their suppliers. These products appear side-by-side with their competitors and are often priced lower. Yet few people argue Safeway should be broken up or that Costco should divest itself of its Kirkland products, for the simple reason that these practices are valued by consumers.
In a widely cited article, antitrust attorney Lina Khan criticized Amazon for competing with Quidsi, an online company that sold baby products, by lowing its prices on diapers and other baby products. She alleges Amazon lowered its prices by 30 percent in order to force Quidsi out of business. While lower prices might have hurt Quidsi, they clearly benefited consumers, at least in the short run. In a deal reviewed by the Federal Trade Commission, Amazon eventually purchased Quidsi in 2011. Yet Amazon continued to face strong competition from other suppliers including Pampers and Huggies as well as marketplaces such as getdyper.com and boxed.com, and of course other major retailers that sold online and offline.
The Quidsi case is worth examining in some detail. Khan basically alleges that Amazon sold diapers at a loss in order to encroach on Quidsi’s business. She may be right about the pricing but wrong about the intended target. As Kristian Stout and Alec Stapp of the International Center for Law & Economics relate, Quidsi’s founders also sold diapers below costs. Why? Because they saw Walmart and Target using diapers as a loss leader in order to build a relationship with new mothers who hopefully would then purchase a lot more from the store.
It might look like Amazon eventually won the battle when it was able to purchase Quidsi for $545 million. But even after this purchase it faced strong competition that limited its ability to raise prices. According to Khan, in 2016 Amazon had 43 percent of online diaper sales, while Walmart and Target had 23 percent and 18 percent respectively. Signficantly more sales were made off-line. But the story does not end there. As Jeff Eisenach of the American Enterprise Institute points out, in April, 2017 Amazon shut down Quidsi, including Diapers.com. Meanwhile the founders of Quidsi used the proceeds from the sale to create another online retail company, Jet.com, which Walmart purchased for $3.3 billion. In one of the most often cited examples of unfair competition, it is difficult to see any harm to either consumers or the founders of the target company. Meanwhile Amazon suffered large losses on both its sale of diapers and its acquisition of Quidsi.
In the vast majority of cases, however, the interests of a platform coincide with those of the businesses that sell over it: to increase sales. The more a given business sells to customers, the more the platform realizes in commissions. Thus platforms have an incentive to help small businesses reach new customers and increase their sales. This is the same practice as in grocery stores, which take a markup on the products they sell and, by offering a wide variety, attract more customers. Customers are better off with a few large marketplaces competing vigorously than with many smaller less efficient shops offering limited selections.
Another issue arises when platforms have access to extensive information about customers and suppliers. The fear here is that they will either use it to compete against suppliers or, by denying it to sellers, they will limit their ability to grow. Again, supermarkets offer a good guide. Walmart, Costco and other big supermarkets collect lots of data about the performance of both their own brands and those of others. But we do not regulate whether they should share this information with suppliers like Proctor and Gamble. Some stores may want to share consumer information in order to attract more suppliers, which in turn will attract consumers. Other stores may want to build a reputation for protecting consumers’ privacy by not sharing data. This might attract more consumers and therefore more suppliers. In each case the store has a strong incentive to attract and keep participants on both sides of the market.
The same dynamic operates when it comes to bundling services. Any seller of a hardware or software product faces a decision of how much to include in the offering. But this is not new. Automobile companies have long bundled features like radios in new cars. While this might have hurt particular sellers of car radios, it clearly benefited consumers who wanted to be able to buy a car with the radio already installed. We see the same dynamic in the digital era. For example, initially Microsoft Word did not come with an integrated spell checker. Consumers had to buy one separately and install it. But it became clear that spell check was an integral component of any word processing program and so Microsoft, Apple and other providers of word processing systems bundled them into the initial offering.
More recently we see companies bundling offerings in smart phones. For example, once Apple realized that consumers valued a flashlight app in the iPhone, it decided to include it when they are sold, in part because makers of many phones using the Android operating system were doing the same thing. This may have hurt a small app developer that might otherwise offer a similar app. Despite this, the market is resilient enough that iPhone users can still use the Apple store to download a number of other flashlight apps for free. In addition, Apple’s store offers an instant market to millions of other app developers. Antitrust policy is meant to promote competition and encourage productivity, not to protect the interests of existing suppliers, be they large or small.
Internet marketplaces benefit consumers in other ways. By reducing the cost of communication and record keeping large technology firms also likely increase the optimal size of many businesses. This puts pressure on small companies to grow enough to achieve the efficiencies that new technology offers. However, platforms give these companies the option of putting some of these burdens onto platforms, effectively outsourcing and turning them into variable costs.
This move toward efficiency largely benefits consumers, who after all are the primary intended beneficiaries of antitrust policy. Large companies often have lower marginal costs from economies of scale, and therefore lower prices. Buyers often prefer large marketplaces that offer a broad selection. But in order to offer a wide selection, platforms need to be large and deal with many suppliers. Although businesses must pay Walmart and Amazon a share of their revenue in order to access these platforms, they obtain access to a large number of buyers without having to duplicate the platform’s infrastructure.
As a filing by ITIF’s affiliate the Center for Data Innovation recently concluded:
While the kind of competition platforms enable is unique, many of their other economic characteristics—such as catering to two sided markets, the fact other businesses rely on platforms as routes to market, the fact platforms often compete with those same businesses, and the fact platforms control large quantities of valuable data, occur in various sectors, as the above comments illustrate.
It is for these reasons, as well as the great diversity of platforms’ business models, that broad regulations on platform-to-business relations would be ill-advised, because they risk imposing restrictions on business models they are not suited to, which would stifle competition and innovation at the expense of consumers. Policymakers should apply the tools they already have at their disposal for promoting fairness and enforcing the law, and limit regulatory interventions to specific, clearly-defined sectors where they will work as intended, without creating harmful distortions.
Still, some regulators are concerned that platforms may have too much power over smaller companies. A report by the European Commission worried that:
Where business models of entire ecosystems of [small and medium enterprises] are dependent on access to a small number of online platforms, or where platforms have access to datasets of unprecedented size, new asymmetries may be created. In such situations, some suppliers to platforms can be disproportionately exposed to potentially unfair trading practices, even in the absence of established dominance of a platform.
Fears about market dominance have led to several unwise policy recommendations. I would like to briefly explain why this is so. First, a number of commentators recommend breaking up the largest Internet companies. Rather than refer to any concrete harm to consumers, these advocates concentrate on the alleged dangers that large companies pose not only to small business, but to our political and social systems. Yet breaking up companies is extremely difficult; it takes years and seldom leads to better market performance. It is also doubtful that courts would approve such drastic moves. Moreover, doing so would almost surely reduce innovation, productivity and consumer welfare. For example, it is hard to not believe that breaking up Amazon into its regular website and Amazon Marketplace would harm consumers, who would then have to go to two separate websites to shop for the same item, thus reducing competition and choice.
Others advocate for legislation that would dictate specific rules for how platforms operate, such as search fairness, data collection or sharing, and removal from the platform. While platforms should publish clear rules and enforce them evenly, dictating the specifics of those rules would harm the market. Each platform already has a strong incentive to satisfy both buyers and sellers. Platforms often differ on how to do this, but the competition between platforms benefits users. For example, some platforms resist sharing data about their search algorithms because they do not want suppliers to game the system. Government officials, lobbied by companies on one side or another, are unlikely to find the best solutions. Moreover, existing antitrust laws already limit truly unfair trade practices.
Extensive limits on data collection and use are also unwise and would hurt small businesses. Consumer data is often helpful in increasing both sales and customer satisfaction. Although much of this data is available from other sources or has a limited shelf life, small businesses are often the least able to get access to it. Where platforms decide to use this data to help suppliers find potential customers, the government should allow it, subject to reasonable privacy rules.
Advocates also complain that, because many platforms get to see what people are looking for, they have an unfair advantage. This is most often expressed about Amazon, which sometimes competes against other companies by offering its own products. But again, large bricks and mortar retailers also get to see what people are buying, and they can and do use this to identify products to sell. Overall, this kind of competition benefits consumers. Again, the law should not protect companies, even ones with a nice story to tell, from the effects of legitimate competition.
Finally, some foreign countries allege that platforms are not paying their fair share of taxes. France recently responded by enacting a Digital Services Tax narrowly focused on the largest U.S. Internet companies. But recent studies show digital companies often face a higher marginal tax burden than large companies in other industries. Moreover, the French Digital Service Tax clearly violates its trade agreements with other countries. It also violates the spirit of existing bilateral tax treaties.
So what can policymakers do to strengthen the relationship between Internet companies and small business? One solution is to enhance rules governing fairness and transparency. Platforms should publish their rules for dealing with both buyers and sellers and should apply those rules in a transparent, even-handed manner. A recent example is the United Kingdom’s open banking initiative and the European Union’s second Payment Services Directive, which will boost competition in financial services by allowing banking consumers to share their data with third party service providers using open application programming interfaces developed by industry. But the scope for doing this is likely to be limited because multi-sided platforms already have a strong incentive to cater to the needs of each side of the market. Perceived unfairness to suppliers not only drives them away, it also makes the platform less attractive to buyers, who now have access to fewer sellers.
Second, regulators can look for barriers to entry, whether natural or artificial, and try to remove them, thus encouraging more platforms to enter the market to serve the diverse needs of small businesses and their customers, both current and potential. Determining the best ways to add value is difficult and often subject to failure, but the general goal of offering services that allow business owners to devote more effort to improving their products and customer service remains an important market opportunity.
Finally, in some cases a limited number of firms outside the technology industry have created an exclusive access to particular datasets that they use to exploit market power. Examples are the Multiple Listing Services in real estate, bank restrictions on financial data aggregators that show customers how to reduce fees, and flight availability and pricing in the airline industry. Some health care providers, and their technology providers, have also been found to be blocking legitimate access to patient data to protect their own economic interests. In each case companies are blocking consumer data without a legitimate business purpose. Regulators should require them to make the data widely available so that consumers can benefit from enhanced competition.
A particular platform may remain successful for a prolonged period if it successfully adapts to new technology and business models. But this is unlikely to systemically harm small businesses who have a number of potential routes to their customers. Platforms already compete for their business and shutting off access would threaten the main source of revenue for many platforms: selling ads targeted at the consumers that visit their sites or encouraging consumers to buy their own products. And again, if such a market position is abused, existing antitrust remedies are adequate to restore competition. For example, if a platform applied stricter criteria or higher fees to products that compete with its own or arbitrarily sold one company sales data about its rivals, we believe existing authority would allow regulators to challenge it.