The Death of US Manufacturing has been Greatly Exaggerated

I often get comments from my friends that they look to my blog to get an optomistic viewpoint on the economy given all the "doom and gloom" that pervades most economic and political discourse these days. Seeing as that I call my base case view of the current economic situation a "Rounded Bottom", I figured optimism is clearly a relative concept. That said, I generally do believe in a self-correcting system, and so therefore counsel against panic and dispair. Tough times make people gloomy, and gloomy people call for more radical action than is generally needed. Often times, it is the very government action designed to goose the economy in times of distress that sows the seeds of the next economic downturn (see housing market, the).

It is in times like these that declinist theories roam the land. While I don't deny that the United States has faced a nasty cyclical decline in the housing market that may have led to a secular downshift in consumer spending and debt accumulation, I don't necessarily view that as a long-run negative. Consumerism, while beneficial to one's near term standard of living, can be carried too far if it becomes a debt-fuelled bacchanal that diverts resources from other productivity-enhancing investments like business equipment and infrastructure.

Exhibit A to most declinists is the supposed decline in American manufacturing. This seems like a no-brainer given our huge trade deficit, the large declines in manufacturing jobs and the visible industrial ruins in former manufacturing hubs like Detroit, Cleveland, Pittsburgh, Philadelphia and Baltimore. A closer look at the numbers, however, tells a much different story.

In October 2011, after one of the nastiest recessions and slowest recoveries in modern history, the United States produced $3.3 trillion worth of manufactured goods on an annualized basis (in 2005 dollars). To put this into perspective, the entire GDP of Germany in 2010 was only $3.3 trillion (in 2010 dollars, no less). China's GDP is only $5.9 trillion. The U.S. is thus by far the largest manufacturer in the world. While production is down about 5% from the $3.5 trillion produced at the end of 2007, and it is up 5% from what was produced in 2000. It is fully 80% higher, on an inflation-adjusted basis, than the manufacturing production in 1979, when manufacturing was still the centerpiece of the U.S. economy.

Look at the chart below (click to enlarge):

Slide1
For comparison purposes, it is important to compare similar points in the business cycle. In the case of manufacturing production and capacity, it is most meaningful to compare cycle peak to cycle peak. I have also included the mid cycle break points (1984 and 1995) where the currency and policy regimes changed somewhat.

The things to notice in the chart above are (1) the long term decline in the peak capacity utilizations (meaning that capacity has increased faster than production); (2) the huge increase in manufacturing production per employee; (3) the decline in manufacturing employment as a percent of total employment from 24% in 1973 to 9% in 2011; and (4) the more modest decline of manufactured final goods production as a percent of GDP from 24% in 1973 to 19% in 2011.

Given the huge gains in technology, finance, professional services, leisure and hospitality, retail, healthcare and education since 1973, the fact that manufacturing has only declined by 4 percenage points from the 1973 peak to the 2007 peak is quite surprising.

In terms of employment, we have to look at manuacturing as the new agriculture. It is a hugely efficient, highly capitalized economic sector that just doesn't employ that many people anymore. I fully believe in supporting manufacturing as a way to increase national wealth, but any politician who tells you that we can bolster the middle class with tons of new manufacturing jobs is out of touch with reality.

Nominal numbers don't provide as much context as relative numbers, however. The following chart translates the numbers above into annual growth rates by business cycle (click to enlarge):

Slide2
The interesting thing about the chart of above is how the numbers vary from business cycle to business cycle, but that they are pretty stable over the long run. The first thing that jumps out is the huge increase in manufacturing employee productivity since the early 1990s, particularly relative to employees as a whole (as defined by real GDP per employee). The second thing that jumps out is the huge surge in manufacturing capacity (5.4% annualized from 1995 to 2000) in the late 1990s bubble boom. Since capacity growth was so far above trend in the late 1990s, the relatively low levels of business investment of the 2000s is not surprising. The huge surge in the capital-to-labor ratio, combined with advances in information technology, has helped create a large increase in per employee productivity…also not terribly surprising yet highly beneficial for the long run health of the economy.

If we look at the longer 16-17 year Kuznets cycle of long-lived investment (consisting of three business cycles as I define them), we see consistent results. Real GDP grew about 3% p.a. peak-to-peak in both the 1973-1989 cycle and the 1989-2007 cycle. Manufacturing production grew 2.4% in the second cycle versus 2.1% in the first, even though manufacturing was viewed to be in decline in the 1990s and 2000s. Capacity grew 2.6% per annum in both cycles. The big difference between the performance under the two cycles is in employee productivity, which grew 3.9% per annum during the second cycle and only 2.5% in the first. Manufacturing productivity growth far exceeded productivity growth in the economy as whole over both cycles.

The good news is that were are now pretty close to having worked off the excesses of the late 1990s. For that reason, I expect the front end of the current Kuznets cycle to produce a powerful resurgence in business investment, which we should see accelerate over the next 6-8 years (with perhaps one recession occuring dueing that time). We can already see it in the numbers. While everyone is focused on the travails of the housing market, the U.S. economy has gradually become a lean and mean manufacturing powerhouse.

More on American Manufacturing

In my previous post on US trade deficits and manufacturing productivity, I discussed how US manufacturing has actually thrived, even in the last three decades while manufacturing employment was plunging. 

Such numbers might prompt one to ask how it can be, when their are abandoned, crumbling factories all around us, particularly in the old industrial cities of the Northeast and Midwest.  There are two answers.  One is that alot of manufacturing activity has moved from the unionised North to the right-to-work South and West, where factories are allowed to be run with fewer employees that cost less.

Another is that the US has been abandoning lower tech, mostly labor-intensive, manufacturing and reallocating capital to higher tech areas.  While this is no consolation to former textile workers in the US, it's the natural evolution of things.

This chart from the Economist illustrates it nicely:

Economist mfg chart

Most advanced nations have been shifting away from low tech manufacturing and such manufacturing is being picked up by developing nations like China. In the US and Britain, where capital has been allocated the most freely, the trade balance is most positive in high tech manufacturing.  South Korea has also made great advances in high tech. 

I think the most interesting aspect of this chart is that Germany and Japan, the two largest advanced economies besides the US, have huge surpluses in mid-high tech but deficits in high tech.  I would argue the reason for this is that Germany and Japan have been among the most aggressive in promoting an industrial policy to build up the old mass market manufacturing economy (exemplified by autos) and have financial systems dominated by commercial banks.  They have very conservative systems that were great at building up the post World War II economy, but lack the venture capital/ IPO/ public stock/ high yield debt/ private equity culture that, for all its faults, relentlessly reallocates capital to the highest-return investments in innovative sectors.  They are also the two developed countries most manifestly in danger of being in secular decline.

I realize that sometime the US system gets its capital allocation very wrong, like with the recent housing bubble.  But I would like to bring up the fact that housing is the sector most heavily promoted by our government.  This "industrial policy" was successful for years, but eventually it got taken way too far and will be difficult to unwind, given all of the powerful constituencies that are now invested in promoting real estate speculation.

Housing in the US and the long term stagnation of Japan and Germany are warnings to those that would promote the heavy involvement of the US government in the economy to protect declining industries and to conduct industrial policy in growth sectors beyond the early R&D phase.

The truth about US trade deficits and US manufacturing

One of my central economic tenets is that the total debt burden of the United States is too high. I blame economic policies that have encouraged excess consumption and housing investment, funded by money borrowed from abroad and by excess leverage in the financial system. Borrowing money from abroad in excess of money invested abroad creates a capital account surplus for the United States. If the US runs a capital account surplus, it must run a corresponding current account deficit to balance its accounts. The biggest component of a current account deficit is the trade deficit. I therefore believe that the best long-term fix for our problems is to adopt a set of policies that for a period of time encourages the US to run trade surpluses and a capital account deficit. In other words, we should sell more US goods abroad, while we increase savings and repay our foreign creditors.

My four point plan for running a trade surplus

  1. Maintain a stable currency and discourage the building of foreign dollar reserves;
  2. Improve incentives / remove disincentives for investing abroad;
  3. Remove / scale back policies that encourage borrowing and consuming at the expense of saving and investing; and
  4. Use policy to improve our terms of trade on energy

Right now point (1) is happening, whether on purpose or by accident. I don't expect (2) to happen anytime soon, and in fact policy seems to be running in the opposite direction. Obama talks like he believes in (3), but his policies also seem to be running in the opposite direction, with higher taxes on capital and high levels of federal borrowing. On the other hand, Obama is clearly a believer in point (4).

The composition of the US trade deficit

In the first quarter of 2009, the trade deficit was equal to 2.4% of nominal GDP (subject to revision). From 1980 to 2008, the trade deficit has averaged 3.1% of GDP.

61% of the deficit in Q1 2009 was due to our trade deficit in petroleum products. Since 1980, petroleum has accounted for 45% of our trade deficit, and the petroleum deficit has been about 1-2% of GDP. This one product is the "low hanging fruit" when it comes to targeting out trade deficit.

Not surprisingly, the US typically runs a surplus in services. In Q1 2009 the trade surplus in services was 1.1% of GDP, which is near the average level since 1990.

The level of trade deficit in goods excluding petroleum has averaged about 2% of GDP. In the goods sector, the US runs at about a balance in foods, industrial supplies and capital goods. In capital goods, the US runs a surplus in civilian aircraft (Boeing) and a deficit in everything else, including computer equipment. The bulk of the non-petroleum trade deficit is tied to autos and consumer goods.

It's not surprising that the US would run a deficit in consumer goods, many of which are labor-intensive like textiles, toys and such. The US should be able to run a trade balance or even a surplus in autos, which is actually a very productive sector in the US, but one in which we are up against rather unfair trade practices abroad. The US should also be running a bigger surplus in capital goods and non-petroleum industrial supplies.

The truth about US manufacturing

It has been so long since the US has run trade surpluses that it seems we have come to believe that it would be impossible to do so again. It is commonly believed that the US has basically stopped "making stuff". Nothing could be farther from the truth. The US is actually the largest manufacturer in the world by far and in most years is the largest exporter in the world. Also, it is important to note that the US's economy is absolutely massive relative to any other country in the world, and that trade (both imports and exports) is a relatively small part of our economy compared with even large economies like China, Germany and Japan. Also, the bulk of our trade deficits since 1980 have coincided (slightly lagged) with two periods of a very strong dollar and high real interest rates, which leads to poorer terms of trade for US manufacturers and increased investment flows from abroad. If we focus on maintaining a stable dollar rather than a "strong dollar", most of the trade deficit problem goes away.

But isn't manufacturing in the United States in decline? While manufacturing employment is certainly in long-term decline, manufacturing production continued to climb right up through 2007. The recent plunge in production has brought US manufacturing back to levels last seen in 1999. Manufacturing employment, on the other hand, is back to levels last seen around the start of World War II. The difference between these two numbers is the increase in manufacturing productivity. Before the recent downturn, we were producing roughly 7.5 times what we were producing in 1948 with roughly the same amount of manufacturing workers, working out to real productivity growth of 3.3% per year.

Manufacturing production and employment:

During the same timeframe, manufacturing capacity has been growing at 3.5% per year, so capacity utilization has fallen.

Thus, when this downturn ends, the US has plenty of scope to have its manufacturing production return to its trend line, particularly in high tech equipment and other durable goods.

The key takeaways are:

Don't confuse manufacturing employment with production. Manufacturing employment since the early 1970s has been analogous to agricultural employment in the late 1800s and early 1900s, declining even while production boomed.

US durable goods manufacturing is perfectly competitive. Capital-intensive durable goods manufacturing in the US could thrive if the US government halted the currency policies that were unduly punishing it.