>This report does not acknowledge the effect of contracting out
>abroad, which is
>part of offshoring.
Here's the full text. They try to capture some of this through service imports.
Doug
----
US Economics Analyst
Offshoring: Where Have All The Jobs Gone?
We estimate that relocation of US production to overseas affiliates accounts for 300,000-500,000 job losses over the past three years. This trend historically has been confined to the factory sector, but cheap telecommunications and huge potential savings will encourage offshoring of service functions to grow rapidly.
Even so, international relocation is quite small compared to the size of the US economy - worth about 0.1% of employment per year. It is one factor behind the current "jobless recovery," but hardly the only one.
A firm choosing to cut costs by international rather than domestic outsourcing will reduce demand for US labor, boost imports, and (on the margin) create downward pressure on the dollar. In the short term, industry leaders may be able to capture some of their cost savings in the form of higher profits; eventually, rivals should compete away these gains.
We estimate that US producers have relocated 300,000-500,000 jobs to overseas affiliates over the past three years. International relocation historically has been confined to the factory sector, but it is now broadening to include service occupations that use cheap telecommunications to leverage lower foreign labor costs. While foreign investment creates many jobs in the United States, relocations intended to lower production costs are overwhemingly outflows. International relocation is small compared to the size of the US economy - about 0.1% of employment per year. It is one factor behind the current "jobless recovery," but hardly the only one. International outsourcing [footnote: 1We avoid the generic term "outsourcing" to describe this trend, as it refers to contracting-out of business processes to third parties without regard to their location. We use "international outsourcing" or "offshoring" to describe the process generally, and "relocation" when a US firm retains ownership by outsourcing to a foreign affiliate.] - whether for goods or services - matters because it affects the US labor market, the macroeconomy, and the balance of trade. A firm choosing to cut costs by international rather than domestic outsourcing will reduce demand for US labor, boost imports, and (on the margin) create downward pressure on the dollar. In the short term, industry leaders may be able to capture some of their cost savings in the form of higher profits; eventually, rivals should compete away these gains and reduce the overall US price level.
The "Giant Sucking Sound"
The manufacturing sector traditionally has been the most sensitive to international trade and the relocation of production. It has been hardest hit in the current downturn, losing almost three million jobs in the past three years - one-sixth of the sector's employment. Manufacturing employment peaked at 19 million in 1979. Exhibit 1 shows the trend since then: stability or slight increases during expansions, sharp declines during and after recessions. This pattern, completely at odds with the steady growth of the service sector, helps to explain why the decline in US manufacturing periodically takes center stage as a political and economic issue: first in the early 80s, then in the 1992 election, and again now.
The change in manufacturing employment has three root causes, two of which are illustrated in Exhibit 2.
_ A shift of consumption towards services. Consumers now spend about 40% of their income on manufactured goods, versus more than half in the late 1970s.
_ Productivity improvement. Capital investment and better organization have enabled companies to produce more goods with fewer employees.
_ Growth in imports. Manufactured imports have risen from about 25% of goods consumption in the late 1970s to roughly 40% today.
Ascertaining the level of international outsourcing activity is easier said than done. Production shifts may be intended to serve foreign markets rather than the US, and firms may use foreign contractors rather than their own international affiliates (though the latter two have similar impacts on the US economy).
We attempt to reconcile data from several sources: relocation announcements, foreign employment of US companies, and changes in import flows. The most direct approach is simply to sum up company relocation announcements. A study by the US-China Security Review Commission in late 2000 to early 2001 found annual relocation of 70,000 jobs to China and 60,000 to Mexico. 2 These two countries appear to capture the vast majority of relocations. The timing of the sample may bias the figure upward, since employment fell considerably in 2001 and announcements presumably lead relocations by several months. On the other hand, a significant number of moves probably go unreported.
Another approach is to look at foreign employment by US multinationals. US nonbank firms had about 7.5 million overseas employees in 1999, up roughly 2.5 million over the previous decade. Based on data from the Bureau of Economic Analysis, about 15% of these affiliates' sales were back to the United States. If we take a leap and assume a similar breakdown of employment, and a one-for-one displacement of US workers by foreign workers, this implies job losses of about 40,000 per year. This lower figure is not so surprising, as it is based on older data and explicitly excludes offshoring to third-party contractors.
Finally, we can look at import trends to gauge the magnitude of relocation. Exhibit 3 shows the annual increase in imports coming from foreign affiliates of US companies. Goods production per worker in the US was approximately $150,000 in 2000; assuming this crude ratio holds for the displacement effect of imports, the import data suggest a transfer of 70,000- 100,000 jobs per year in the late 1990s. Note that this only includes jobs in US-owned subsidiaries, not those contracted out to foreign suppliers. Each of these approaches has its flaws, but collectively they present a reasonably consistent picture. We estimate that relocation within US multinational firms has accounted for the transfer of at least 100,000 US manufacturing jobs per year, or roughly 300,000-500,000 over the past three years - even more when third-party foreign contractors are included. China appears to be the largest beneficiary, as it represents the largest US bilateral trade deficit, and US firms have invested over $14 billion there since the business cycle peak in 2000.
Outsourcing affects the economy in several ways. As foreign production comes on line, imports increase, putting downward pressure on US goods prices. Plant closings in the United States reduce domestic labor demand, and drive up unemployment in the short term.
This shift, and the resulting flow of imports, creates an imbalance that can be resolved in one of two ways. First and most likely, the dollar can depreciate, improving the relative competitiveness of US domestic production. Second, over the much longer term, the shift in labor demand will shrink the gap between foreign and US wages, decreasing the incentive for further relocations.
What's Different This Time
In terms of its depth, the current downturn in employment has precedent in the post-World War II period (refer back to Exhibit 1 on pg. 4). In the three years leading up to December 1982, the economy lost 2.6 million jobs; in the three years to August 2003 the total was 2.7 million. The strength of the dollar contributed to the depth of these two troughs: Its trade-weighted value in each case was 10%-20% higher than in the 1990-1991 downturn, as illustrated in Exhibit 4. In the slowdown of the early 1990s, the weaker dollar may have helped domestic manufacturers to avoid deeper labor cuts.
What distinguishes this slump in employment is its length. It took about six months for employment to show significant growth after the trough of the 1981- 1982 recession, and about a year after the 1990-1991 recession (two years for manufacturing). This time around, it is almost two years and counting. A study by the Federal Reserve Bank of New York3 suggests that the lengthy period of slack reflects a larger structural component of unemployment. In particular, most industries that lost jobs during 2000 have continued to do so since, and vice versa. This is a different pattern than past recessions, where many cyclical manufacturing industries rehired employees on the demand rebound, and it is consistent with the notion that production is being relocated outside of the United States.
For the long term health of the US economy, the biggest concern is what happens to those workers who lose their jobs. The theoretical benefits from trade assume that laid-off workers find some type of employment; otherwise overall US income would fall. Clearly, most workers do find new employment. Even after three years of steady job losses, the US unemployment rate and labor participation rate are at almost exactly the same levels as in 1979, at the dawn of the decline in manufacturing. This masks considerable churn - over ten million workers were displaced from their jobs because of plant or shift closings in the 1999-2001 period - and associated economic hardship. In addition, the relative quality and pay level of the replacement jobs is less clear.
The Fate of Your Barber Technological and regulatory changes in recent years have expanded the opportunity for trade in service industries such as financial services, software, and telecommunications. At present imports of such business services remain small - about 6% of total imports and less than 1% of GDP in 2002.
Could services trade and outsourcing eventually become as significant as manufacturing? While activities such as telephone customer service have already begun the move offshore, others (the classic example is haircuts) seem inherently unsuited to international trade. One projection (by Forrester Research) suggests that over three million service jobs could move offshore by 2015. We think this number is somewhat conservative, based on two alternative estimates. In the first, we estimated the share of jobs that could be relocated abroad on a sector-by-sector basis, based on conversations with industry experts. In the second, we took the service industry as a whole and estimated the share of each job type (occupation) that could be offshored. For example, in 2001 there were 2.5 million US jobs in customer service and telemarketing occupations, many of which are vulnerable to international outsourcing. Depending on the aggressiveness of the assumptions, our two approaches suggested that international outsourcing could affect up to six million jobs over the next decade.
On the surface, this suggests that the long term trend could be as significant as that in manufacturing. But it is still early days - the most recent data show only $15 billion of service imports per year from foreign affiliates of US companies, suggesting cumulative offshoring of less than 200,000 employees. The pace of change depends on at least three factors:
_ Magnitude of potential savings. Estimates range widely and depend on the specific occupational tasks, but savings of 50%-80% in labor costs are feasible in many cases. While services outsourcing eventually may boost foreign wages, the incentives will remain high for many years.
_ Management constraints. It's hard to imagine a large company relocating half of its employees in only one or two years; capturing the full savings is likely to take five years or longer.
_ Capacity of the major offshore destinations. Most of the jobs to be offshored involve some level of English and/or technical skill. India, the largest offshore services destination, produces roughly 750,000 college graduates per year. Even if two-thirds of these are hired by foreign firms, this puts a ceiling on the pace of change. On balance, we think it is possible for services offshoring to ramp up to a few hundred thousand jobs per year over the next two to three years. This may result in a temporary boost to profitability for firms in these sectors, although that is likely to be competed away over time by equally low-cost rivals.
- Andrew Tilton