Why American Manufacturing Moved Offshore

Why American Manufacturing Moved Offshore

LEARN / REINDUSTRIALIZATION

Why American Manufacturing Moved Offshore

Offshoring followed a sequence rather than a single cause, and the order matters more than the list. Cost differentials made the move worth considering, logistics made it practical, capital markets made it attractive, and trade policy removed the remaining risk. Reversing a sequence requires knowing which step is still reversible.

Key points

The sequence

Explanations of offshoring usually take the form of a list: low foreign wages, permissive trade agreements, corporate greed, weak enforcement. Each item has some support and the list format obscures the thing that matters, which is that these factors arrived in an order and depended on one another. Low foreign wages had existed for as long as anyone had measured them, and produced no great migration of production until several other conditions changed alongside them.

A wage differential is worth acting on only if the goods can be moved cheaply, the quality can be verified at distance, the capital can be raised for the transition, and the political risk of depending on a foreign supplier is tolerable. Each of those became true at a different time, and offshoring accelerated as each one resolved.

Logistics: the enabling condition

Containerized shipping collapsed the cost of moving manufactured goods across oceans, and collapsed the variance of that cost along with it. Predictable transit converts a distant supplier from a gamble into a line on a schedule. Without it, a wage differential of any size is offset by inventory carried against uncertainty, and firms hold production close to demand because they cannot plan otherwise.

Logistics is properly an enabling condition rather than a cause. It did not make offshoring attractive; it made offshoring possible, which is a different claim. It is also the step least likely to be reversed by any policy, because the ships, the cranes and the port capacity are already built and their operating cost falls with use.

Capital markets: the pressure

A factory is a large, illiquid, long-lived asset that depresses return on assets while it is being paid for. A supply contract with an overseas producer is none of those things. For a firm judged on quarterly earnings and on capital efficiency, converting owned production into purchased components improves nearly every ratio that gets reported, and does so immediately rather than over the life of the asset.

This pressure operated independently of any trade agreement and independently of wage differentials large enough to matter on their own. It explains why offshoring proceeded even in cases where the landed cost advantage was thin, and it explains the asset-light strategies that treated manufacturing as a commodity input to be purchased rather than a capability to be held.

Institute analysisCapital market pressure is the driver least addressed by current industrial policy. Subsidies change the cost of building a plant. They do not change how a plant is treated on a balance sheet or how a management team is evaluated for owning one. A firm that builds with public money and is still judged on return on assets faces the same incentive to divest the asset once construction incentives lapse.

Trade policy: the timing question

Trade policy is the driver most often named and the one whose timing fits least well. The North American Free Trade Agreement took effect on January 1, 1994. Congress granted China permanent normal trade relations through the U.S.-China Relations Act of 2000, and China acceded to the World Trade Organization in December 2001. Manufacturing employment had peaked in June 1979 and had been falling through recessions for fifteen years before the first of those dates.

Measurement noteChronology constrains causation without settling it. A decline that began before an agreement cannot have been caused by that agreement, and can still have been deepened, accelerated, or made permanent by it. Studies attributing employment losses to trade exposure and studies attributing them to productivity growth generally use different periods, different industry groupings, and different counterfactuals, which is why they reach different totals from the same underlying data.

What trade policy did unambiguously change was risk. A tariff schedule subject to annual congressional review is a different planning environment from a permanent one, and permanence is what allows a firm to commit capital to a foreign supply base rather than hedge against it. The effect operated on investment decisions as much as on prices.

Which steps are reversible

The practical value of separating these drivers is that they do not respond to the same instruments. Containerized logistics is not reversible and there is no reason to want it reversed. Wage differentials narrow slowly through foreign development and cannot be closed by American policy. Trade policy is directly adjustable, which is why it absorbs most of the attention, and it acts on relative prices rather than on the physical constraints that limit what can be built.

Capital market treatment of productive assets sits between these. It is neither a law of nature nor a statute, and it responds to accounting standards, procurement structures, and the terms on which long-lived assets are financed. Of the four drivers, it receives the least policy attention and is the one where the mechanism is most clearly domestic.

Common misconceptions

That a single agreement caused it. The employment peak precedes every agreement usually blamed for it, which means the agreements shaped the path of a decline already in progress rather than initiating one. Arguments that begin in 1994 or 2001 are describing an acceleration and calling it a start.

That low foreign wages are a sufficient explanation. They were available for decades before offshoring scaled, and a standing condition cannot explain a change in behavior unless something else changed alongside it. The useful question is what changed in the interval, which points at logistics, capital markets and political risk rather than at the wage gap itself.

That reversing the policy reverses the outcome. Tariffs alter relative prices. They do not by themselves produce transformers, machine tools, trained welders, or interconnection capacity, and where those are the binding constraint a price change moves nothing.

What the evidence says, and where it is contested

ContestedThe relative weight of trade exposure and productivity growth in explaining manufacturing employment losses is genuinely unsettled in the economic literature, and the disagreement is methodological. Estimates vary with the period selected, the level of industry detail, and the counterfactual assumed about what would have happened absent the trade shock. Both mechanisms are present in the data. The dispute concerns how much of the total each accounts for, and no study design has settled it.

Related Institute research

When Did American Deindustrialization Begin?
Why the starting date depends on which series you count.

Is American Manufacturing Actually Declining?
Three positions on what the employment fall demonstrates.

Reshoring Nitrile Gloves: America’s Cheapest Industrial Vulnerability
What bringing a single product back actually costs.

Sources

Reference entry maintained by the Institute for American Manufacturing & Technology. Figures are drawn from primary sources and cited above. Where the Institute states a position rather than a finding, it is marked as such.

Private Professional Network

Institute for American Manufacturing & Technology

Join the IAMT Network. Connect with researchers, policymakers, and industry leaders working on American manufacturing, energy, and technology policy.

Apply for Membership →