The Jones Act Doesn't Work Like You Think

Autarky and Monopoly

The Jones Act is a common enemy among anyone who is interested in low-hanging-fruit policy changes, but it isn’t universal. Some of the policy wonks at the Niskanen center favor more conservative reforms to the act, and some usually libertarian-leaning economists like Brian Albrecht and Josh Hendrickson express some doubts over criticisms or even explicitly defend the Jones Act. These aren’t shallow or self-interested endorsements, they are cases made by careful thinkers. Are there actually good reasons to support the Jones Act?

The Case for The Jones Act

Sophisticated supporters of the Jones Act go beyond the base appeals to the national glory of supporting American trucking and point out an externality that the Act is attempting to address: military defense. Hendrickson and the Niskanen center claim that the intention and benefit of the act is to subsidize the shipbuilding industry so that there are ships and crews ready in times of war. This externality argument here is clear enough. There is a benefit to the US having access to extra ships in times of crisis that commercial shippers do not get paid for. I haven’t seen anyone make an attempt to calculate how large this benefit is or claim that the Jones Act gets close to the correct amount of subsidy to address it, but at least the optimal subsidy is non-zero.

Lets look at some graphs to understand this better. The pro Jones Act case boils down to the claim that when an importing country passes a tariff on a good, domestic production of that good increases and producer surplus goes up. Since the Jones act completely bans foreign built ships from cabotage, we can model it as a prohibitive tariff that is large enough to make all trade unprofitable. Then standard economic theory predicts an increase in the quantity produced domestically.

Under free trade the price is kept low at the red line by foreign competition. Only the few most efficient suppliers can profit under this price so consumers buy most of their ships from other countries which is why domestic demand exceeds domestic supply. Without trade, domestic supply and demand must be equal. Under the trade ban more ships are produced domestically but fewer ships are used overall. This is the directionally correct response to a positive externality from domestic shipbuilding. Notice also that producer surplus increases. Since they are addressing the positive externality by producing more ships, it makes sense to reward them more than pure competition would.

The Empirics

All of this is standard and well supported economic theory. But little of the empirical evidence seems to fit the story laid out above. Merchant ship production in the United States actually decreased after the passage of the Jones act.

WW1 was a boom for US shipbuilding. Source here and data hosted online here.

The causal effect of the Jones Act is not well identified by this graph but it is certainly not clear that the Jones Act succeeded in increasing domestic shipbuilding.

The domestic ocean shipping market in the US remains miniscule. There are 91 Jones Act compliant ships for the entire United States. Norway builds 5 times the tonnage of the US in absolute terms and over 200 times the tonnage per dollar of GDP. More than three quarters of all shipyards that built new vessels have closed since 1983 and only 4 shipyards that build ocean-going cargo ships remain in the US. This is hardly the picture of an industry whose output we are supposed to be significantly increasing.

It is theoretically possible that the counterfactual US shipbuilding industry is even smaller without the Jones Act but this just doesn’t pass the sniff test. There’s no reason to expect that the US would be a worldwide runt in shipbuilding. You might think that high labor costs and regulations would disqualify a large shipbuilding industry but the top 20 ship exporting nations includes Germany, The Netherlands, Finland, Norway, Japan, and South Korea. The US has the capital, technology, and demand necessary to create a globally competitive shipbuilding industry but we aren’t using it, despite protection from foreign competitors who have already levered up their economies of scale.

There is one piece of evidence which fits with the predictions of standard trade theory: increased producer surplus. Markups of shipbuilders in the US are significantly higher than in other shipbuilding nations.

A Better Model

Two of the three main predictions of the standard trade model are confirmed by the data: higher prices and producer surplus. But do we really believe that one or two ships a year is US shipbuilding at its peak when Norway does 37 and Japan 370? Is there another model that can explain decreased quantities and increased producer surplus following a trade ban?

One assumption of the standard trade model is that, even without international trade, markets are perfectly competitive. Lets return to the our graphs of prohibitive tariffs from above, but this time lets relax the assumption of domestic perfect competition and consider what happens when transitioning to autarky means creating a monopoly.

If the transition from free trade to no trade creates a domestic monopoly then it creates exactly the type of situation we seem to be in with domestic shipbuilding: a protected and highly profitable industry that is producing far below its actual capacity.

To move beyond the static equilibrium in the graph above to a more dynamic picture of how this would arise, lets consider a hypothetical. Imagine that, inspired by the success of the Jones Act, congress decides to expand the domestic manufacturing requirements to air travel. This would leave Boeing as the only currently existing company who could supply planes for domestic flights.

In the short run, almost all of the airlines who used Airbus now want to buy from Boeing, so the demand for Boeing’s planes increases. But the marginal cost of airplane construction is a lumpy step function. If Boeing was producing near the maximum capacity of their current plants already, the price increase from the trade ban needs to pay for the huge upfront cost of a whole new plant before any increase in quantity happens. Depending on the size of the initial demand increase, it might not be profitable for Boeing to increase their quantity supplied at all. At the very least it would take time for them to adjust. The huge minimum-efficient-scale of airplane manufacturing means that incumbent firms can absorb large price increases without increasing their quantity supplied or facing new competition.

If the marginal cost and demand curves for airplanes looks something like this graph then we a domestic monopoly would be stable after a trade ban. The competitive equilibrium after the trade ban is where domestic supply (which is Boeing’s marginal cost in the short run) equals demand. That point has higher price and quantity than under free trade, but not much higher. In particular it’s not enough for Boeing, or a competitor, to afford building a whole new factory.

As time goes on, the price increase of domestic air travel will spill over into demand for substitutes like road and rail. Since these substitutes have increasing returns to scale, they will drain the initial demand spike for Boeing making investments in new capacity even less appealing. As the economy settles into its new equilibrium Boeing can change its pricing strategy. Instead of having to cut their markups close to marginal cost to compete with Airbus, they can take advantage of their captive market and cut quantity supplied to raise the unit price.

Consumers respond by flying less, investing in substitutes for domestic flights, and focusing on upkeep of existing planes. Boeing is enriched, the domestic air fleet decays, and non-contiguous parts of the US languish under raised shipping costs and hobbled tourism. Sound familiar?

This model can explain what we observe in the shipbuilding market under the Jones Act. Large fixed costs make the quantity increase in response to increased demand for American built ocean-liners muted and slow. In the meantime, investment flows into substitutes like road, rail, and air. Increasing returns to scale for substitutes accelerate this transition. As the infrastructure and technology serving substitutes improves, the initial demand spike for American ships dwindles and few long term investments are made in the industry. Eventually, only a few firms are left serving the only the most inelastic uses for domestic ocean shipping. These firms enjoy large markups and profits.

This is exactly what we observe. The US is a world leader in rail, air, and road shipping but is a laggard in domestic ocean cargo. Quantities are tiny, producers are few, and markups are large.

The Jones act succeeds in subsidizing incumbents in the US shipping industry but it does not address the military defense externality. This is because it subsidizes the water cabotage industry by protecting monopoly power. Do not mistake increased producer surplus for subsidizing a positive externality. Increasing the quantity produced is the only way to fix a positive externality. Monopoly profits make producers better off, but they do so at the expense of quantity produced which is what we actually care about.

Conclusion

The externality argument that Jones Act supporters make is probably directionally correct. Standard trade theory predicts that a prohibitive tariff will raise domestic production and increase producer surplus. US shipbuilding has double the markups of the rest of the world so we’re confident that producer surplus increased. But the transition to autarky for already concentrated domestic markets fosters monopoly power and producer surplus is funded by quantity restriction. This is the exact opposite of what we want for a positive externality.