Evan Soltas
May 22, 2012

The Coming Services Disruption

I am convinced that a massive disruption is coming to the services industry in the United States within the next decade. By "disruption," I mean major productivity gains which will compel an equally large reallocation of labor, capital, and technological resources in that sector.

You might as well re-title this post "Why Nursing Is No Safe Bet." If you look at the Bureau for Labor Statistics' projections for employment growth between 2010 and 2020, "personal care aides" and "home health aides" hold the first and second spots.

However, I think there's a very real possibility that this won't happen -- that instead, we could in fact regular service-producing jobs could meet the fate of comparable goods-producing jobs. That is, mass creative destruction.


Labor costs in services are outpacing in labor costs in goods industries, whereas goods productivity has soared relative to services productivity -- and the way goods industries began to behave as they do, with such fast productivity growth and so much employment hemorrhaging, emerged from disruptive technological innovations which were introduced, considering the data, in the late 1990s. It looks totally unlike the pattern of employment vs. production growth in the services industry. As if the two exist in completely different technological contexts. That does not seem sustainable, and therefore by Stein's Law, it will stop. The sort of technology which seems likely to emerge, moreover, will not be of limited consequence or factor-neutral -- employment in the services industry appears highly vulnerable to sweeping, skill-biased technology.Such disruptive innovation would dramatically increase the marginal productivity of capital (MPK) in services industries. Formally stated, when this happens, it will push out the labor-capital isoquant and cause firms to reallocate their resources to a new point on their isocost lines until, where r is the cost of capital and W is the wage:


To write that in English, firms will take the resources they currently have employed in labor and shift them to capital -- read: they will fire people and buy equipment and technology -- until an additional purchase of capital would do them no better, or would be as good a deal as, an additional hire.

The incentives for such a change are huge, as cost inflation in services has been extreme -- see my post on Baumol's cost disease -- creating a large incentive for investment in technology, because a technology which can reduce the marginal cost of production to something a little bit more reasonable will create substantial economic profits for such firms in the short run.

Another way to think about this: This is what Ford's Michigan plant looked like in 1910. This is what a similar Suzuki plant looks like today. This is what an elementary school classroom looked like in 1910. This is what it looks like today.

Where are the greatest marginal returns on investment in technology likely to be found?