President Trump appears to have a vendetta against Amazon because he’s angry about how the Washington Post, owned by Amazon founder Jeff Bezos, has covered him and his administration. The company filed a complaint last week in the U.S. Court of Federal Claims, saying Trump had improperly interfered to prevent a $10 billion defense contract from being awarded to Amazon; it went to Microsoft instead. In the complaint, Amazon chronicled the connection between the president’s ire about press coverage and his beef about the JEDI defense contract.
As my colleague Jonathan Chait has argued, the president’s practice of using departments of the executive branch to punish people and firms that publish news he doesn’t like is a major scandal undermining the free press and the rule of law. But in addition to the broad impact it has on our institutions, the president’s other major initiative to punish Amazon and Bezos may have significant, direct impacts on consumers.
That is, Trump has repeatedly pressured the U.S. Postal Service to hurt Amazon by raising its package-delivery prices. So far, he has been hemmed in by bureaucratic requirements, binding postal contracts, and (for now) not having his preferred personnel in charge of the post office. But soon, a new U.S. postmaster general will be appointed by the Postal Service Board of Governors, a majority of whose members finally are Trump appointees. If Trump gets his way and forces package-delivery price increases, that could mean not only a higher cost to consumers for Amazon Prime but higher prices for shipping from all vendors, enriching UPS and FedEx (which could simultaneously raise prices and gain market share) at consumers’ expense. Paradoxically, Trump’s demand that the post office raise prices could even hurt the post office’s own finances.
When Trump has called for higher package-delivery prices, even Trump-skeptical press outlets have tended to give the president more credence than he deserves. After all, when he says the post office loses about $1.50 on every Amazon package it delivers, he’s not just pulling that number out of thin air; he’s citing a 2017 research note from Citibank. Right?
Unfortunately, I think I’m the only person who actually dug into the research note. As I wrote last year for Business Insider, the Citi note doesn’t say what the president says it says, and it contains a serious math error that causes it to greatly overstate the point it intended to make. (Remember: Banks can be wrong about things.) Citi makes no convincing argument that the post office is losing money on Amazon.
Laws governing the post office divide its business lines into two categories: market-dominant products, like first-class mail, on which it faces a legal mandate to provide universal service and enjoys a monopoly; and competitive products, like package delivery, on which it competes with private firms such as UPS and FedEx. The post office is legally obligated to price its products competitively. Services like package delivery have to be priced roughly high enough to cover both the marginal costs of delivering each package and an appropriate share of the post office’s fixed costs, such as central administration and certain real-estate expenses.
Until this year, that “appropriate share” was set by regulation at 5.5 percent. But in its 2017 note, Citi (repeating an argument UPS had made to the Postal Regulatory Commission) said package delivery had grown and grown as a share of post-office business, so a more appropriate share would be 24.6 percent. The problem is Citi did not realize the post office was already far exceeding its legal obligations under the appropriate-share rule. As of 2017, competitive-product sales were covering 23.2 percent of the post office’s unallocated general costs, meaning competitive-product prices would have had to rise by only 11 cents per piece to meet Citi’s proposed appropriate-share requirement, not the $1.46 that became the headline number out of the Citi note.
Last year, a Citi research analyst admitted to me that the bank had erred in arriving at the $1.46 figure, yet the president has relied on it, further misrepresenting it as an estimate of losses specifically on Amazon deliveries (and sometimes rounding up to $1.50). In fact, USPS pricing offered to Amazon and other major shippers is proprietary, so Citi could not and did not estimate a profit or loss per piece on Amazon deliveries, only on competitive products overall — including packages sent by other shippers and non-package products such as Express Mail.
The post office, for its part, views Amazon deliveries as an important piece of its strategy to minimize losses at a difficult time when it must adapt to falling first-class-mail volumes. It is difficult to do even marginal cost analyses on an entity like the post office, which cannot adapt to changing market conditions in the same ways a private company can. Because of the universal-service mandate and political constraints that make it difficult to lay off employees or shed excess real-estate holdings, the post office must be able to expand in a business (package delivery) that allows it to redeploy trucks, buildings, and employees that would otherwise be operating below capacity because of its declining market-dominant businesses. A rule forcing the post office to price packages higher, losing market share to private competitors, would lead to more unused capacity in the existing postal trucks, not to a right-sizing of the post office.
So far, the post office’s argument that its current package pricing is good business has won out. Early this year, the PRC raised the appropriate-share requirement to 8.8 percent and said it will change the requirement in future years in response to market conditions. The post office opposed this move from the fixed 5.5 percent floor, wanting to maintain maximum flexibility in its pricing; but in practice, this was not enough of an increase to materially impact either the USPS pricing model or Amazon’s shipping costs.
But a task force report ordered by the Treasury Department last year makes clear the administration’s desire for higher package-delivery prices to avoid “potential market distortions” from prices that are too low for private firms to compete with. And starting next year, the president is likely to finally have leadership that he likes in place at the post office. In October, Postmaster General Megan Brennan announced her retirement, effective in January. A majority of the sitting members of the Board of Governors, which will choose her replacement, are Trump appointees.
The president wants to hit Bezos where it hurts, and undermining Amazon’s key competitive advantage in retail — reliable, attractively priced, fast shipping — would certainly be a way to do it. But if Trump succeeds, he’ll be hitting consumers in the process. According to eMarketer, about half of U.S. households have at least one member with an Amazon Prime membership, spending an average of $1,400 a year on products from the company. Adding about $1.50 to the delivery cost for each Amazon package would have a material affect on the Prime membership fee (currently $119 a year), product prices, or both.