No one can accuse the Teamsters of thinking small. In mid-August, the union and its allies gathered outside New York City Hall to rally for the Delivery Protection Act, a sweeping bill that would effectively ban subcontracting in the last-mile-delivery industry. The bill is a top priority for the Teamsters, who have faced growing competition from independent delivery service providers (DSPs). Though these DSPs are small individually, together they employ thousands of drivers—more than five thousand in New York City alone. And in recent years, they have grown fast to meet the ever-rising demand for home delivery—demand that might otherwise have been fulfilled by card-carrying Teamsters. So the union has a problem, and its solution is straightforward: ban DSPs.

The ban, however, will probably fail. The reason is basic economics. DSPs have proliferated in part because of advancing technology: smartphones, GPS, and online marketplaces have made it easier to contract, track, and coordinate with multiple firms. In economic terms, the technologies have cut transaction costs. And when transaction costs fall, economic theory tells us that firms will do less internally. Instead, they will buy more services from independent suppliers.

That’s exactly what’s happening with delivery services. Because transaction costs are lower, firms can more easily find and contract with delivery specialists. These DSPs are more efficient because they focus only on delivery. They’re specialized. So if firms can outsource to them without too much cost, they will.

That’s why the ban won’t work. It effectively tries to roll back the technological clock. And fighting against technology is a losing game: governments have tried it before, and they have almost always failed—though often not without causing a lot of unnecessary pain. One hopes that New York City can learn from those mistakes without first having to repeat them.

Old Ideas in New Bottles

The Delivery Protection Act is a big swing. It aims to reverse decades of growth and change in the delivery industry. The industry was once dominated by a few big players, including the United States Postal Service. But in recent years, as ecommerce has taken off, it has diversified. New demand has drawn in small startups, some with no more than a handful of drivers. And these small firms have found willing partners in big retailers and ecommerce companies eager to diversify their delivery channels. So as delivery volume has continued to rise, a growing percentage of it has been absorbed by small DSPs.

The Act tries to roll that change back. In broad terms, it requires certain last-mile-delivery companies to employ all the workers who perform “core services.” Core services include both warehouse and delivery operations. For example, if a logistics company wants to deliver from its own warehouses, it has to employ its own drivers. But DSPs are by definition not employees: they are independent firms that specialize in delivery. So the Act effectively makes their business models illegal.

This strategy isn’t new. It showed up last year in another New York City law, the Safe Hotels Act. Among other things, that law required large hotels to employ the workers who perform their core services. There, core services were defined narrowly to exclude things like food service and parking. And the hotels were allowed to keep some existing third-party arrangements in place. But the basic idea was the same: companies had to employ the workers who did their core work.

This idea makes a lot of sense for at least one group–labor unions. When a union organizes a firm, that firm usually pays higher labor costs. The firm can then be undercut by nonunion competitors, risking the jobs of the union’s new members. A union can combat that risk by organizing any rival firms, too. But organizing is costly, especially when workers are spread across multiple firms. Pulling them all into one firm makes organizing easier. And that makes defensive organizing a more viable strategy. 

Transaction Costs as Market Destiny

But in the long run, the core-services idea is swimming against the tide. The trend toward subcontracting isn’t an accident, and it isn’t limited to the delivery industry: it is economy-wide, and it is driven by technological change. Technology has made it easier for firms to spin off ancillary functions to small providers. And when firms can spin off functions profitably, economic theory tells us that they will.

Start with the theory. In The Nature of the Firm, a seminal work on organizational economics, Ronald Coase explained that a firm can provide a good or service in two basic ways: make or buy. That is, it can make a thing itself or buy the thing on the market. Which one the firm chooses will depend mostly on market prices. In a vacuum, the market should always be able to provide the thing more cheaply: someone will specialize in making the thing and so will produce it at a lower cost. There should be no reason to make most things in house.

But of course, firms don’t make decisions in a vacuum. They have to consider not only the cost of the thing itself, but also the cost of buying the thing. They have to find a supplier, negotiate a price, and monitor the supplier’s performance. In Coase’s formulation, these costs are “transaction costs,” and they mostly dictate how big a firm will get. The firm will grow to the point that producing one more thing in-house costs less than the market price, transaction costs included. Everything else, it will buy.

Critically, this make-buy line isn’t static. It moves as transaction costs go up and down. And one thing that can drive transaction costs down is technology. As technology makes it easier to find and coordinate with suppliers, people can work more efficiently across firm lines. For example, smartphones make it easier to communicate with contractors. GPS makes it easier to monitor contractors’ services. And digital-payment technology makes everything easier to buy. Together, these technologies have driven down the costs of outsourcing to the market. And as transaction costs have fallen, firms have found it economical to spin off more and more tasks.

Lessons Learned and Learned Again

That background offers an important lens for the Delivery Protection Act. The Act aims to block delivery outsourcing. But the shift to more outsourcing is driven largely by technology, and technological change is hard to reverse by fiat. People have tried before, and they have mostly failed.

A few examples stand out. Maybe the most embarrassing is the Red Flag Law, an effort by British lawmakers to slow the development of the automobile. In the late 19th century, steam-powered autos started showing up on Britain’s roads. Their appearance threatened incumbent industries, such as railroads and stagecoaches. These industries lobbied Parliament, which responded by shackling autos with regulation. For example, it required each auto to be staffed with three operators: two in the cab, one walking out front waving a red flag. It also restricted top auto speed—four mph in the country, two in the city.

These restrictions worked as intended: development in the nascent industry ground to a halt. But they also had unintended consequences. While the British auto industry stagnated, its German and French rivals boomed. Foreign automakers quickly gained ground on their British rivals, with even British engineers fleeing overseas to work on the new technology. Parliament eventually relented and repealed the law, but not before losing three decades’ worth of ground.

Similar mistakes have played out in the United States. For example, in the 1990s, the U.S. government tried to stop “strong” encryption technology from spreading through the private sector. It tried to block the technology with multiple policy levers, including classifying the technology as a “munition” under international arms-control rules. But the restrictions proved so unpopular and ineffectual that they were abandoned by the end of the decade. And what followed was the modern internet marketplace: secure online transactions, safe email, and a robust international exchange of ideas.

Lawmakers have also tried, and failed, to restrict technology-driven independent contracting. In 2019, California lawmakers passed a strict worker-classification law aimed at the “gig economy.” Their idea was to force app-based platforms to treat rideshare and delivery workers as employees. But the effort was so unpopular that voters rejected the change by ballot measure, carving out app-based drivers by a margin of almost six to four. So today, app-based delivery and rideshare workers are still independent contractors, and they are still providing the same independent services in the Golden State.

These examples could be multiplied. But that doesn’t mean they will be heeded. As politics too often reminds us, the lessons of history and economics are among the hardest to learn. Lawmakers are motivated less by the historical record than by political exigencies. They respond to well-organized interest groups, who want protection whatever the cost. These lawmakers are unlikely to be moved by the failure of earlier legislative efforts, and they are even less likely to be moved by dry economics. They are more than willing to trade sound policy for good headlines. 

But headlines are ephemeral. And in the long term, market forces eventually win out. Technology is hard to bottle up; and as technology pushes the make-buy line down, firms are likely to follow. Core-services laws might slow that trend, but they are unlikely to reverse it.