Vertical Integration

Over the past 100+ years, industrial capitalism has seen waves of vertical integration and outsourcing. During the first two-thirds of the twentieth century, integration was all the rage; firms encompassed a large part of the production process, housing suppliers, suppliers of suppliers and so on under a common corporate roof. The classic example was the large, integrated auto company, with Ford famously going so far at one point that it tried to establish its own rubber plantation in Brazil. But despite continued horizontal integration (“conglomerate” firms), stages of the production process were progressively spun off after about 1970, so that supply chains of at least nominally independent firms gradually replaced in-house sources of materials, services and components. That has seemed to be the norm until the last decade or so, but now there are signs that integration is coming back, at least through partnerships and cooperation agreements, if not direct corporate reabsorption. These trends have important effects on economics and politics, both nationally and globally, and it would be helpful to understand them.

But it’s not so easy. In fact, there are several theories that have been offered to explain the rise and decline of vertical integration, and we have little basis for weighing one against the other. Worse, it’s common for proponents of a given theory to view it as the only one, so the question of relative weights never arises. This is why it could be a useful service to simply enumerate them.

1. The costs and limitations of contracts. Among economists, this, the Coase-Williamson theory, remains hegemonic. Ronald Coase, a brilliant theorist of the intersection of law and economics, posed the question this way: if markets are so wonderful—the most efficient way to coordinate any set of economic activities—why do we see firms bringing intermediate products and processes together within the same legal entity and coordinating them administratively? The answer, he thought, had to be that the legal mechanism on which markets rely, the formulation and enforcement of contracts, has costs that in many cases outweigh the advantage of having multiple parties competing for your business. Oliver Williamson took this a step further, proposing that the main cost with using the market is that you make yourself vulnerable to the self-interested stratagems of those you do business with. They may deceive you in drawing up or executing contracts, or they may take advantage of your dependence on them once it’s been established. This makes the command system of in-house management, with its powers of surveillance and control, look good in many circumstances.

2. The economies of administration. This view is most associated with Alfred Chandler, who argued that in many cases it is simply cheaper and more efficient to coordinate a range of related activities through direct management than to try to bring the pieces together by buying and selling on the market. That’s especially true when the timing of these activities is important: making sure productions schedules are aligned so part A is always paired with part B, and so on. He put the production systems of several large, integrated firms under the microscope to see how technical and market strategy considerations dictated which suppliers would be brought in-house and which could remain independent.

3. The dynamic economies of coordination. This seems obvious, but as far as I know, I’m the only one to spell out this aspect of the problem. It’s an application of a broader “theory of the firm”: firms exist because productive activities are typically interactive, and the “right” way to perform one of them depends on how the others are performed. If this is the case, there’s an argument for combining decision-making over all of them together. This consideration gains much more force in the context of innovation, where successful identification and implementation of new options depends not only on coordination but the information suppliers have about how their products might be produced and not just how they currently are. You could think of operating through the market instead of a planning structure as limiting, although this effect is reduced as it becomes possible to specify the characteristics of components (specs) with greater precision, adjust them more quickly—even continuously—and transmit to interested parties a wider range of potential products than just the few currently on the market. The IT revolution of the past 30 or so years has had just this effect, so you might see it as a partial explanation for the outsourcing trend—but then how to explain the apparent reverse of this trend recently?

4. Tax, labor and regulatory arbitrage. As the barriers, technical and logistical as well as economic/legal, were falling over the final decades of the twentieth century, it became increasingly possible to source much lower-cost inputs, especially labor, from countries that organized themselves as export platforms. China in particular served this role. Similarly, it became profitable to outsource many intermediate products and processes to countries that used their lack of regulation or low taxes as competitive advantages. This is the “race to the bottom” thesis, and it surely played at least some role in encouraging firms to substitute external supply chains for internal integration. China’s further development, especially since around 2010, which has reduced its labor cost advantage and generated increasingly detailed and extensive regulatory constraints, has moderated some of these incentives. We haven’t seen the longer-term effects of the most recent retreat from trade liberalization yet, however.

5. Dual economy factors. The profitability of different investments depends to a large extent on how competitive the markets are for their outputs. As the technical ability of firms to outsource increased, they also increasing took advantage of differential profitability potentials. The activities core to establishing a firm’s brand, which target market share in oligopolistic settings, can expect to yield higher returns; those that produce goods sold in competitive markets (commodities) will yield less. This is a justification for spinning off the latter and concentrating your capital in the former—ROI is king. There is much empirical evidence for increasing profit rate differentials and corresponding wage inequalities (rent-sharing). Since there is no evidence, as far as I know, that this landscape has been changing in the direction of a less dualistic structure, this is not a theory that predicts a reverse move toward vertical integration.

These are the theories I’m aware of. If there are others out there, let me know. And what we most need right now are case studies that show concretely which factors are most salient or changing in importance in specific instances.