Skip to main content

What Trump Has Wrong ‐ and Right ‐ About Amazon

Irwin M. Stelzer

President Trump’s use of the presidency to pursue a private vendetta with Amazon is appalling. But that doesn’t mean policymakers shouldn’t take a close look at the internet giant’s power and competitive tactics.

Trump doesn’t like Jeff Bezos because he doesn’t like the Washington Post. Just as he doesn’t like Time Warner because he doesn’t like (to put it mildly) CNN. But he can’t do much about either the Post or CNN; something called the First Amendment annoyingly stands between him and the vengeance he seeks. And he can’t do much about Bezos, personally, in part because the Amazon founder is surrounded by an impenetrable wall of money that protects him from just about anything Trump can throw at him. Yes, the president can tweet away billions of the value of Amazon shares, but that’s petty cash to Bezos, only harms smaller shareholders by cutting into the value of their pensions and 401(k) plans.

Claiming that Amazon doesn’t pay sales taxes won’t work because Amazon does, at least on its own merchandise. Claiming that the Post Office is subsidizing Amazon won’t work either, because it isn’t. Or if it is, the place to find out is in a proceeding before the Postal Regulatory Commission, which has the thankless task of allocating the cost to be borne by each class of users benefitting as these couriers complete the sometimes swift completion of their appointed rounds. The Commission says the postal service is charging Amazon the same rate as it charges others bulk customers, and if it can wring more out of Amazon in the next round of negotiations, so be it.

Bezos has been too busy building one of the most consequential businesses of our time to write a book about “the art of the deal.” So he might be disadvantaged in a face-off with the author of a work on that subject. But I rather doubt it.

Start with this. The use by the president of the United States of the power of his office and pulpit to settle personal/political scores is unseemly and reveals—if any such further revelation is necessary—that Trump is a fledgling caudillo, unable to differentiate between the personal and matters of proper concern to the nation’s chief executive. His admiration for, or jealousy of, Vladimir Putin knows no bounds.

But the impropriety of Trump’s actions does not mean that Amazon’s success doesn’t create real policy problems that ought to be considered and, after consideration, either acted upon or laid to rest.

That consideration should take two forms. The first is the more traditional examination of whether Amazon possesses monopoly power as that term is generally understood. Translated into the jargon of competition policy, that means measuring Amazon’s share of some relevant market to see if that share seems so large as to give it substantial power over prices to the detriment of consumers.

The quick answer to this query is that Amazon possesses no such market share. About 10 percent of all retail business is done online, and best estimates are that Amazon accounts for about half of that, or about 5 percent of all retail sales, with sales of third-party merchandise accounting for more than half of that 5 percent. That’s a lot of sales, but hardly a monopolist’s share.

Moreover, Amazon is facing increased attacks on its online stronghold from traditional retailers. Its (far larger) retail rival, Walmart, has just dropped “Stores” from its corporate name in recognition of its belated (but serious, and so far successful) thrust into online retailing. Bezos has brilliantly constructed a way of doing business that is forcing others to imitate Amazon’s higher level of customer service. And so far there is no apparent barrier to competitors doing so.

But that’s not the end of the tale. A second set of considerations requires policymakers to examine two further questions. The first is whether Amazon’s market share in specific product categories is sufficient to give it market power that might not be in consumers’ long-term interests, and is not reflected in its low share of the total retail market.

Here the evidence is unclear. Yes, Amazon has cut the retail price of books and music. But lower prices now are not necessarily in the interests of consumers if they harm competition in the long run. What we do not know, and an antitrust investigation might discover, is whether these prices are predatory—to be raised once competition is more or less eliminated. (We could also test whether there is a relationship between Amazon’s margins on products and its share of the markets in those products.)

The second area of possible inquiry involves Amazon’s pricing and other business policies to determine whether it uses financial muscle, rather than mere efficiency, to nip would-be competitors in the bud.

In the long run a competitive economy depends on innovation, and innovation depends on the ability of newcomers to raise capital, either by initial public offerings of shares in the new enterprises, or by persuading venture capitalists to back them. The latter are notably hard-headed realists. If they believe that an entrenched incumbent can snuff out a potential competitor merely be announcing an intention to enter the market he has targeted, they will look for other opportunities. Distributors will be unlikely to add new products to their lines. Retailers will not allocate precious shelf space to them. Venture capitalists and retail investors will at the very least raise the price they ask for the use of their capital to reflect the enhanced risk, and more likely suggest to the newcomer that completion of his doctoral dissertation or a job with the entrenched incumbent is his best option.

It is notable that so many startups that might have blossomed into competitors of Facebook and Google ended up selling their companies to the incumbents, who were prepared to pay prices that reflect their desire to have the new firm inside rather than outside of their tents. And that Amazon was able to pummel the price of Blue Apron’s IPO merely by announcing that it was considering selling the dinner-making kits that Blue Apron had on offer.

So much for the mere economics of the policy problem. I say “mere” because the antitrust laws were never intended to be aimed solely at maximizing efficiency and achieving the lowest possible prices for consumers.

When Trump said “Towns, cities and states are being hurt—many jobs being lost,” he had a kernel of an idea wrapped in either a lie or a misconception.Total jobs are not being lost, as the obscene competition among various venues for Amazon’s second headquarters (and its 50,000 jobs) shows. Amazon’s gigantic, efficient infrastructure has created tens of thousands more jobs than traditional retailers have lost.

What the president might have had in mind after studying the origins of our antitrust laws is that consumer welfare was not and ought not be the sole goal of antitrust policy. Our antitrust laws also have social and political goals, not least among the latter the diffusion of private power—particularly important in this day when private power travels to K Street, where it is transformed into public policy. This coexistence of economic and social goals is not peculiar to the antitrust laws. Trade policy must take into account what we now see, thanks in part to Trump and in part to new scholarship, as the devastating effect on communities of the competition of imported steel and other products, a high cost not reflected in what consumers pay for imported steel.

Some examples: Tax policy aims not for the most efficient economy, but one that is arguably as efficient as fairness permits. Financial regulation must treat small community banks differently from giant financial institutions, lest we ignore the community-wide effects of the unavailability of credit to small businesses. When we close military bases it surely is appropriate to consider the effect on the stability of entire communities. Regulators of electric rates set rates for street lighting low enough to enable often financially-strapped cities to pay for sufficient lighting to maximize the crime-reducing effect of street lights. Minimum wage laws do not aim to keep prices to consumers at their lowest possible level; they aim to offset imperfections in labor markets and mitigate the socially undesirable effects of even perfectly functioning markets.

In all matters, efficiency is tempered with considerations of fairness and the effects of policy on parties other than the first-level consumers. So, too, it would it be sensible public policy to weigh the effect of Amazon on more than the net cost of goods to consumers. That is the truth hidden in Trump’s tweet-rant against Amazon.

Easier said than done, you might respond. Fairness, or equity, is a concept so elastic that politicians can stretch it to cover just about any policy they believe will prevent an involuntary move to the private sector. And you would be right. But better attempted than ignored, and better to accept imperfect solutions that assure a system fair enough to sustain wide support, and cost-efficient enough to sustain high and improving material prosperity. One reason for our political paralysis is that conservatives attempt to set policy that maximizes efficiency, ignoring the voters’ sense of what is fair, while liberals aim for fairness with no concern for the effect of their policies on the costs of those programs and on the ability of the economy to bear those costs.

Which brings us back to Trump and Amazon. The crippling of Amazon would be a high price to pay for damaging Jeff Bezos, and by extension the Washington Post. And neither would be a proper goal of policy in a nation of laws. But allowing Amazon to continue many of its present policies without examination—acquisitions, pre-announcements to make it difficult for entrants to raise capital, pricing that might be predatory— might damage competition.

And surely it is worth knowing whether Amazon’s delivery infrastructure gives it sufficient power to deny access to producers of goods with which it competes or hopes to compete.

The proper balance between efficiency and equity will not be arrived at by studying tweets. Or by unthinking expansion of an already overweening regulatory state. But it might be arrived, dare I say it, by establishing a new national commission to study our competition policy, which was conceived and largely shaped before the emergence of internet giants and same-day free delivery. The old-time antitrust religion might just prove up to the task of preserving an innovative, dynamic, competitive economy. But then again, it might not.

Related Articles

The Politics of Inflation

Irwin M. Stelzer

Expect inflation and a devalued dollar...

Continue Reading

China’s Black Box Superiority

Bruno Maçães

The Chinese approach to AI has implications for Chinese society that could herald a new age...

Continue Reading

China Reports Weak Manufacturing Growth Amid Trade War Woes

Michael Pillsbury

Michael Pillsbury on the trade war between the U.S. and China....

Watch Now