Monday, March 03, 2008

NAFTA Bashing: Pure Politics Unsupported by Data

"We're going to take a hard look at NAFTA, which has a lot of problems," Hillary Clinton said. "We're going to renegotiate it."

"If you travel through Ohio and you travel through communities in my home state of Illinois, you will see entire cities that have been devastated as a consequence of trade agreements that were not adequately structured to make sure that U.S. workers have a fair deal," said Senator Obama.

Comment: As the chart above shows, more U.S. jobs have been created in the 14-year period after NAFTA was passed than in the 14-year period before NAFTA. Payroll data from the BLS, via FRED, are available here, and civilian employment data here.


At 3/03/2008 10:35 AM, Blogger Bill Combs said...

I think in your "comment" you really mean to say the opposite of what you say.

At 3/03/2008 10:59 AM, Anonymous Machiavelli999 said...

US Unemployment rate day NAFTA is enacted = 6.3%

Today (in a 'recession') = 4.9%

At 3/03/2008 11:37 AM, Blogger spencer said...

Is this how they teach you do do econmic analysis at GMU?

You should compare the percent increases rather then the absolute numbers.

Take the payroll number you use, for example. You show a gain of 21.4 million in the first period and 25.6 million in the second.

The gain in the first period is 24% while the gain in the second period is 13% .

That means that job gains after NAFTA were nearly cut in half.

At 3/03/2008 1:15 PM, Blogger Walt G. said...


I analyzed NAFTA using percentages obtained from BLS data. I used the percentage increase of civil labor force participation pre-NAFTA from Jan 1980 (106,562) to December 1993 (129,941) which is a 22% increase and post-NAFTA from Jan 1994 (130,596) to Dec 2007 (153,866) which is an 18% increase. I was expecting to find that there were fewer jobs for more people on a percentage-adjusted basis, but the data did not support my hypothesis.

I would like to blame NAFTA for the decline in the auto industry, that’s where my bread is buttered, but my past research has shown that productivity gains are the main culprit in manufacturing job loss. Blaming NAFTA does not solve any problems. Politicians’ pandering to the public does not solve any problems. Adapting to the change in the job market--and a willingness to move to available work as my parents did--will.

At 3/03/2008 2:19 PM, Blogger spencer said...

i happen to agree that NAFTA is being blamed for a lot of other problems --
just like Wal-Mart is blamed for a lot of problems.

But that still does not excuse sloppy economic analysis. Comparing the number of jobs created pre and post NAFTA rather then the growth rates is just plain poor economic analysis.
That is my problem with the post not any of the points made or conclusions reached.

At 3/03/2008 5:01 PM, OpenID reality-hammer said...

Don't forget that outsourcing can change the same job (say, drafting) from a "manufacturing" sector job to a "service" sector job.

I worked a technical job (software engineer) for 12 years through a service firm, for example.

At 3/04/2008 5:05 PM, Anonymous Steve said...

What are the inflation-adjusted average and median wages, including benefits, of the jobs created and lost during the two periods?


At 3/05/2008 7:13 AM, Blogger Walt G. said...


The data you seek are on the Bureau of Labor Statistics website. You can’t generalize because there are over 800 occupational groups with different rates of wage increases during those two time periods. As a rule, those with more education and higher skills were rewarded for their efforts over those with less education and fewer skills.

At 3/05/2008 10:28 AM, Anonymous Steve said...

Thanks for the response. My point is that comparing absolute numbers or growth rates of jobs pre-Nafta and after is pretty meaningless without more information on the jobs being lost vs. the jobs being gained. A $50K/yr manufacturing job with full benefits, for example, is not exactly comparable to an $8/hr service job without benefits.

I haven't checked BLS info recently, but I imagine that there are aggregate income data available.

At 3/05/2008 11:40 AM, Blogger Walt G. said...

Pre-NAFTA was a 7.03% decrease in pay and after NAFTA was a 9.54% increase in pay using data from BLS from below. This is for production workers, and accounts for inflation.

Series Id: CES0500000032
Seasonally Adjusted
Super Sector: Total private
Industry: Total private

Jan 1980 $8.11
Dec 1993 $7.54

Jan 1994 $7.55
Dec 2007 $8.27

At 3/05/2008 12:19 PM, Anonymous Steve said...

This tells us how much an average worker in the same employment category made across two time periods. It still tells us nothing about the aggregate employment picture. Again, remuneration for manufacturing job is not comparable to that for a job stocking store shelves or flipping burgers.

What is needed are numbers demonstrating changes in relative levels of employment across employment categories AND average salaries, to achieve a weighted average of income compensation for the entire U.S. population across time. Median figures are also necessary to account for non-normal distribution and extreme outliers.

Then we can have a meaningful picture of how the "average" American worker is doing before and after.

At 3/05/2008 12:56 PM, Blogger Walt G. said...


Then you're back to picking and choosing among the 800 NAICS categories. All the data are there: wages, hours worked, and number of employees in each category.

It sounds like a good research project for someone. I've already got three major projects going on now that are due before May, or I would give it a shot.


Post a Comment

<< Home