Wednesday, May 5, 2004

previous entry | main | next entry | TrackBack (0)

Fun with BLS numbers

The Bureau of Labor Statistics has a Mass Layoff Statistics program, in which firms that lay off 50 or more workers must provide as reason for such a move to the BLS. Those reasons range from automation to product line discontinuation. Offshore outsourcing is not one of the options, but "import competition" and "overseas relocation" are options. So, it's possible to estimate the extent to which offshore outsourcing is respinsible for job destruction via mass layoffs [How do you know that the firms aren't lying to the government?--ed. You don't -- but since the names of the firms are kept strictly confidential, there's no reason for them to lie either]. You can do it too -- just click here to create your own table.

Here are the percentages of jobs lost through mass layoffs because of either import competition or overseas relocation for the last seven years:

1996: 1.78%
1997: 1.87%
1998: 2.10%
1999: 2.50%
2000: 1.82%
2001: 2.88%
2002: 1.89%
2003: 2.41%

Now, these figures do not cover instances when a firm let go less than 50 people, so clearly there's a bias in the data towars multinational corporations over small businesses. That said, these numbers reveal two important facts:

1) Offshore outsourcing is not responsible for a significant percentage of the jobs that have been lost.

2) There is no evidence that offshore outsourcing is responsible for an increasing number of jobs lost over time.

Finally, some have argued that the massive increases in U.S. labor productivity are due to sloppy GDP accounting: "[T]he work done by Indian software firms is being recorded as US economic activity and growth because it's been offshored." If true, this would be a serious measurement error, since the government would be overstating both economic growth and labor productivity

The BLS issued a memo in late March on this very issue back in March that's worth perusing. The highlights:

[W]e have experienced nearly 13 years of faster productivity growth. While a number of explanations have been put forth and to this list some have added measurement issues related to outsourcing and offshoring, any set of explanations should cover not just the last few years, but the entire 13 year period....

Offshoring affects business sector productivity change only through changes in the composition of domestic production and its effect is likely to be small. In manufacturing, the combination of domestic outsourcing and offshoring has contributed about 1.5% per year to sectoral output per hour growth through 1995 but only about 1% per year thereafter and as a result, they do not appear to be an explanation for the productivity speed-up.

This conclusion must be qualified in two ways. First, there is no information on the relative importance of offshoring relative to domestic outsourcing and so it is not known if foreign suppliers have become a growing substitute for domestic suppliers of intermediate inputs. Even if they have, under reasonable assumptions, offshoring appears to explain only a small fraction of the productivity speed-up. Second, not all BLS data extend beyond 2001 and so it cannot be ascertained if there has been a sudden shift in trends. Even if there has, the impact of outsourcing and offshoring on productivity change is likely to be small.

posted by Dan on 05.05.04 at 12:47 PM


Care to comment on this article?

Did I pass the spelling test this time? ( I used short words)

posted by: Dan on 05.05.04 at 12:47 PM [permalink]

From Dan's link:

The total number of jobs and job opportunities is a measure of what employment in trade-related industries would have been if the U.S. trade balance had remained constant (and holding everything else in the economy constant) between 1994 and 2000.

Does "everything else in the economy" include the net foreign investment in the US which balances the increase in the trade deficit? If so, it's something they had no business holding constant.

posted by: Paul Zrimsek on 05.05.04 at 12:47 PM [permalink]

Hmmm... I should make clear that by "holding everything else constant" I take them to mean (based on the preceding note on "Methodology" in the article) that they're leaving everything about the economy unchanged from its historical trajectory, but holding the trade deficit at its 1994 level. If you read them literally, all you'd get would the uninteresting conclusion that if you start out with the 1994 economy and hold the trade deficit and everything else constant, you also end up with the 1994 economy.

posted by: Paul Zrimsek on 05.05.04 at 12:47 PM [permalink]

Dear Dan,

Thank you for both the citation and your thoughtful response. Whatever our differing opinions, you are no light-weight and have my sincere admiration. I have read the memo, and see a lot of non-denial denials and blanket assertions made without proof. I will be looking up the source citations and pulling these contradictions apart looking for clues.

However just as a red glaring warning sign on the surface I will note two things:
(i) The BLS memo did not deny that offshoring could inflate productivity (even significantly) - it simply argued that it didn't
(ii) According to the quote summary above, the effects decellerated after 1995.

"In manufacturing, the combination of domestic outsourcing and offshoring has contributed about 1.5% per year to sectoral output per hour growth through 1995 but only about 1% per year thereafter and as a result, they do not appear to be an explanation for the productivity speed-up."

However this alone begs reason to differ. As it is popularly known, offshoring has in absolute terms and relative to the economy increased as a practice since 1995.

So how can the BLS estimate of output growth from offshoring have decreased after 1995 by 2/3rds???

This is a glaring red warning signal that something is wrong with the above analysis. When BLS admits that offshoring can inflate output, and then suggests that the output inflation has decreased during the very period when offshoring was growing by leaps and bounds then we have a contradiction in the evidence that suggests a need for deeper analysis.

For the record, I don't believe that there is any more of a conspiracy going on that of bureacracies and Administrations of both sides of the aisle being lazy, out of touch, and wanting to make themselves look good.

However even the BLS memo you quote has a contradiction in the evidence that indicates that not all might be what it appears to first glances. Thank you for bringing it to my attention.

posted by: Oldman on 05.05.04 at 12:47 PM [permalink]

Oddly enough, the number of mass layoffs where no reason for the layoff was reported, a continually growing portion of the mass layoff data, has more than doubled in 2003 over 2002. And the 2003 numbers are not complete yet.

The number itself more than doubled in the five years between 1997 and 2002. It is the only continually increasing annual number in the available range of data. For 2003, this mysteriously unexplained portion accounts for 12.1% of the layoff data set; 2.5 times the portion it represented in the 2002 data and 3.5 the portion it represented in the 2001 data.

So why are corporations suddenly so much less willing to indicate why they are laying off workers? Could it be because the reasons are unsavory? Are they perhaps afraid that bringing attention to the reason could cause a negative political backlash against the practice?

I guess we'll never know, as long as corporations are nervously ashamed about putting thousands of Americans out of work for their own pockets.

Useful site Dan, we should all use it in good health. And in poor income.

posted by: Keith Tyler on 05.05.04 at 12:47 PM [permalink]

I would have thought there are at least two reasons why the BLS numbers are upward-biased (making off-shoring even less of an issue than the raw stats suggest).

First 'import competition' is a much broader term than just off-shoring. A lot of this is regular down-sizing in response to competition from other firms that are overseas-based.

Second, as Dan notes, small companies are missed. I would assume they do less off-shoring than large companies, so a large company sample would overstate the general trend.

posted by: Dan Ryan on 05.05.04 at 12:47 PM [permalink]

Post a Comment:


Email Address:



Remember your info?