Several commenters have responded to my recent story (and blog posts) about the decline of U.S. manufacturing by insisting it's no big deal if the manufacturing sector shrinks. The United States will gradually replace current manufacturing-sector work with higher-value-added manufacturing and service-sector work, the argument goes, just as we replaced agriculture with manufacturing during the last century. (About 40 percent of the labor force toiled in agriculture in 1900; that was down to about 2 percent in 2000.) These critics are confident we either won't miss a beat or will get a lot richer in the process.
I think there are at least three big problems with this argument, which I may elaborate on later, but will just briefly lay out here:
1.) U.S manufacturing has major competitive problems; the agricultural sector doesn't--and, so far as I can tell, didn't develop them as it shrank. American farmers consistently produce more than we consume; American manufacturers do not. In 1980, we ran a surplus in agricultural food production of $20.4 billion; in 2008 that was $32.7 billion (though it was significantly smaller as a percentage of the total amount traded). By contrast, in 1980 we ran a trade surplus of $17.4 billion in manufacturing; last year we ran a deficit of $433.2 billion. So while both sectors have become more productive over time, food production stabilized relative to domestic demand even as the sector shrank, while manufacturing continues to decline relative to domestic demand. That means our foreign competitors our eating our lunch in manufacturing; not so in, well, lunch.
2.) The beauty of manufacturing is that wages and productivity aren't necessarily tied to education level. A person with a high school diploma (or less) can make a middle-class living in the manufacturing sector. (See the example of FormFactor in my piece about Ron Bloom.) By contrast, wages and productivity are much more closely tied to education level in the service sector. A person with a high school education or less will generally do very badly in a service-sector job--there are very few service jobs that can provide them with a middle-class living.
Now it would be great if everyone would go to college and be able to thrive in the post-industrial economy. But, in reality, there's always going to be a significant portion of the population that doesn't get beyond high school. Which means that an economy with almost no manufacturing is probably an economy with much greater income inequality. (This obviously wasn't a problem during the transition from agriculture to manufacturing, for the opposite reason.)
3.) What of the idea that we can continue to shed l0w-value-added manufacturing functions and focus on the most sophisticated and innovative parts of the sector? For example, you could outsource production of laptop components and assembly to Asia, but continue to develop and design laptops in the United States. In that case, we might have a smaller manufacturing sector, but a much more profitable one.
Unfortunatately, the evidence suggests it's nearly impossible to thrive at R&D and product-design unless you're also actively involved in the production process, too. (And that's setting aside the income-inequality issue.) To stick with the laptop example, while U.S. manufacturers initially outsourced less sophisticated components and assembly to Asia, laptops are now almost entirely developed and designed in Asia, too. (Apple is really the only exception to this trend.)
As Harvard business professor Gary Pisano explains in this blog item:
To innovate, you need great two-way feedback. You need to transfer knowledge from R&D into production, but you also need to move knowledge from production back to R&D. The act of production creates knowledge about the process and the product design.
Yes, there are some instances where R&D and manufacturing are separable. But these are the exceptions. In the vast majority of high tech products (and even some low-tech products like apparel), knowledge about manufacturing helps you design products and get them to market quickly. What this means is that when manufacturing capabilities migrate from a country, design and R&D capabilities eventually follow. That's exactly what's been happening in many high tech industries in the U.S. over the past 20 years.
Pisano and his colleague Willy Shih, from whom I lifted the laptop example, had a long piece on the topic in the Harvard Business Review this summer. You can peruse it here (though most of it is behind a subscriber wall).