Arnold Kling  

Krugman and Wal-Mart

Strange Comment on Social Secu... The Fear Factor...

According to George Reisman, Paul Krugman is now accusing Wal-Mart of destroying jobs.

Krugman notes that Walmart’s competition against other retailers also destroys the jobs its competitors had offered before being put out of business by its competition. He attempts to show that Walmart has destroyed more jobs in this way than it has created.

Paul used to argued against the people who claim that international trade destroys jobs. He used to argue for the economist's point of view on trade, which is that it is like technological innovation. It increases efficiency, but increased efficiency does not destroy jobs. It changes jobs around, and overall raises the average standard of living.

That same argument would apply to Wal-Mart, and Paul knows it.

I do not have access to the column, so I do not know if Reisman is being fair to Paul or not.

UPDATE: A reader showed me the column. Here is an excerpt:

Because Wal-Mart's big-box stores employ fewer workers per dollar of sales than the smaller stores they replace, overall retail employment surely goes down, not up, when Wal-Mart comes to town. And if the jobs lost come from employers who pay more generously than Wal-Mart does, overall wages will fall when Wal-Mart moves in.

...Maybe it should consider an alternative strategy, such as paying higher wages.

This is not persuasive. Wal-Mart hires low-skilled workers. If you want the wages of low-skilled workers to go up, what you need are more Wal-Marts, so that there is a lot of competition for them. If you want higher-skilled workers to enjoy high wage jobs, what you need is for them to be in highly productive businesses.

If Wal-Mart pays higher wages without competitive pressure to do so, that does not solve anyone's problem. Krugman really knows better. Or used to.

Comments and Sharing

CATEGORIES: Labor Market

COMMENTS (27 to date)
KipEsquire writes:

There's also a secondary effect that is often overlooked regarding Wal-Mart: a significant fraction of its revenue come from small business (mostly via it's Sam's Club unit).

By reducing costs to these businesses (as well as to consumers directly), Wal-Mart creates jobs and raises wages in, e.g., restaurants, bars, contractors, accountants, law firms, dentists, and on and on and on.

This undeniable benefit to small business (and therefore to small business employment) swamps any purported net loss of retail jobs (which is a dubious proposition in and of itself).

Bernard Yomtov writes:

This undeniable benefit to small business (and therefore to small business employment) swamps any purported net loss of retail jobs (which is a dubious proposition in and of itself).

On what evidence is this assertion based?

Rick Shmatz writes:

Here is the article. This guy copies all of the NY Times articles.

anon writes:

I was assigned Krugman's _The Accidental Theorist_ for Bryan Caplan's Econ of Labor class this past semester. I was at first surprised that Caplan assigned this book, then suprised again when I read the required sections and realized that Krugman actually made some, imo, very good defenses of the free market and the efficiecies it yields.

I haven't read this column of Krugman's so I can't comment. I do enjoy reading him bash the Bush admin and his columns about health care.

Taylor writes:

Paul Krugman is a nut. A freakin' nut. And you know what?! I read the bastards' column every Monday and Fridays anyways. Damnit! He's the Bill O'Reilly of the economic left.

James writes:

Krugman as an economist may have made some valuable contributions to the field, but Krugman as columnist is no more an economist than Michael Moore. The NYT knows that his column won't draw as much readership if he says what he must know to be the truth, that some Republican proposals, that some Democrat proposals are bad, that most of the time markets are superior to any alternative, etc. so they keep him on payroll as long as he stays to the party line.

Andrew writes:

But jobs aren't a scarce good. The amount of work it takes to produce a given output depends on how productive we are.

If Walmart makes society more productive, of course that "destroys jobs". It takes less work to get the same production.

But so what? Isn't that a *good* thing, because it let's us move on to more important pursuits in life, like curing disease and writing econ blogs?

Chris writes:
But jobs aren't a scarce good.

But labor is.

spencer writes:

This is the paper Krugman cites in his column.

Neumark, David, Zhang, Junfu and Ciccarella, Stephen M, "The Effects of Wal-Mart on Local Labor Markets" (November 2005). NBER Working Paper No. W11782

I did not read the paper, but the logic advanced by Andrew in the comment above makes a great deal of sense.

Your analogy with trade is not a valid comparison.

International trade does destroy jobs. but the theory does not claim that the same firm that destroys jobs creates new jobs. What theory claims is tht in a dynamic economy the resources freed by imports will be shifted into a new and more productive use. No where does theory say that one firm should play both the job destroyer and job creator role. I have always assumed that it would be different firms doing the job destroying and job creating. In New England we lost shoe manufacturing jobs to the south and abroad. But we did not turn around and create new job making shoes. Rather we created new jobs manufacturing other products or in education, finance, etc.., But there is no reason to think that the firm losing jobs is the firm creating jobs. Actually, there is a wealth of reasons to assume just the opposite. You argument that one firm-- WMT -- is the same as an economy is not a
valid analogy.

If your analogy were true we could blame WMT for the destruction of high paying jobs in the auto industry because displaced auto workers ended up taking a job at WMT that paid less. It is not the fault of WMT that auto employment is falling.
Moreover, productivity in retail is less then in the auto industry, so you can not blame WMT because prevailing wages in retail are less then in auto manufacturing. WMT is paying the prevailing wage in a low wage industry. There is nothing good or bad about that. We would prefer that people work in a higher productivity, higher paying industry, but that is not the fault of WMT that people do not get jobs in higher paying industries.

Consequently, there is no inherent conflict behind the posibility that WMT or any one firm destroys jobs and good economic theory.

Daniel writes:

Thinking of jobs as goods is imho absurd. Jobs are contracts between employers and employees. If there is a good then its the labour. And since the jobs at walmart are supposedly more productive, they should be higher paid than those which Walmart destroyed. I advise you, Andrew, to stop thinking in classical terms, and start thinking macroeconomic, i.e. common-sense

In my view, there is nothing wrong with walmart, since the goods are hopefully cheaper, the consumers are better off (if not the smaller shops would survive), and should the labour remuneration=labour productivity hold for walmart workers, then its good for the society as a whole.

Only problem are the newly unemployed. The might not instantly find new jobs, because they are not enough qualified for other jobs, are distressed or whatever. Another problem could be the quasi-monopolist status of walmart, i.e. walmart raises its prices after the smaller shops are gone.

TJ writes:

Krugman first makes an honest statement saying "On the contrary, there's every reason to believe that as Wal-Mart expands, it destroys at least as many jobs as it creates...". Later he goes on to say that in the retail sector, there has certainly been a job loss. I agree with everything he says, but the way he says it leaves the less astute reader to believe that Paul believes Wal-Mart reduces total employment. It is yet another example of how Paul has left economics and become an ideologue. Its too bad, he was a great economist. The truth is trade neither creates or destroys jobs. It is the number of people willing to work which ultimately determines the number of jobs.

Steve Sailer writes:

What we need is libertarianism in one country.

In other words, if our Southern border remains porous, than the world's poor will continue to pour in, keeping our own poor and working class' wages down.

Shut the border to illegal immigration and let the free market work within America. Or, you can have the border open and increasing inequality within America that will lead to increasing demands for interventions in the economy.

marc writes:

Wal Mart is not to blame for low wages.

They pay low wages because there is an over abundance of unskilled, uneducated illegal immigrants flooding our border from Mexico, willing to work for sub standard wages.

This is an issue supported by the various unions to the detriment of their members, and by politicians in both parties who are more interested in catering to the various special interests who benefit from this policy.

Want to increase the wages of everyone in this country - SHUT DOWN THE BORDERS!!

Paul N writes:

What we need is real libertarianism.

In other words, if our Southern border remains porous, the world's opportunity seekers will continue to pour in, keeping the prices of U.S. goods competitive on the world market, and maintaining the spirit of innovation and progress essential to the U.S.'s success.

Open the border to legal immigration and let the free market work within the world. Or, you can have the border closed to appease scapegoatists - wages of laborers will rise, but prices will too, and the American economy will become stagnant, with the rest of the world suffering as well: the Smoot-Hawley 1930s all over again.

Movie Guy writes:

Arnold Kling -- "It [international trade or technological innovation] changes jobs around, and overall raises the average standard of living."

"If you want the wages of low-skilled workers to go up, what you need are more Wal-Marts, so that there is a lot of competition for them. If you want higher-skilled workers to enjoy high wage jobs, what you need is for them to be in highly productive businesses."

We must be reading different government and industry reports regarding American average standards of living.

Third Quarter 2005, revised
Bureau of Labor Statistics

The seasonally adjusted annual rates of productivity growth in the third quarter were:

5.4 percent in the business sector
4.7 percent in the nonfarm business sector
3.4 percent in manufacturing
3.2 percent in nonfinancial corporations


Hourly compensation includes wages and salaries, supplements, employer contributions to employee benefit plans, and taxes.

Real hourly compensation takes into account changes in consumer prices.

1. Business Sector Productivity

"Real hourly compensation, which takes into account changes in consumer prices, fell 0.8 percent in the third quarter and [fell] 4.0 percent in the second quarter."

2. Nonfarm Business Productivity

"When the rise in consumer prices is taken into account, real hourly compensation declined 1.4 percent in the third quarter of 2005 and [declined] 3.1 percent one quarter earlier."

3. Manufacturing Productivity

"When the increase in consumer prices is taken into account, real hourly compensation for all manufacturing workers fell 1.9 percent in the third quarter."

"Real hourly compensation dropped 0.4 percent rather than rising, after revision", in the second quarter.

4. Nonfinancial Corporations' Productivity

"When the rise in consumer prices is taken into account, real hourly compensation fell 0.6 percent in the third quarter."

"Hourly compensation also was revised down, from 3.7 to 0.8 percent", in the second quarter. Real hourly compensation declined 3.2 percent during the second quarter.

All part of the revised American economy under the advanced global trade model.

P.B. Almeida writes:

Krugman has really really really officially jumped the shark. I am actually a tad embarassed for him when he writes this:

Because Wal-Mart's big-box stores employ fewer workers per dollar of sales than the smaller stores they replace, overall retail employment surely goes down, not up, when Wal-Mart comes to town.

Um, yeah, I guess that would be correct, Paul. Overall retail employement no doubt does decrease, at least temporarily, when Wal-Mart's increased efficiency enters a new market. It is this same increased efficiency that not only benefits shoppers, but also benefits employers and firms in the neighborhood by freeing up labor for more productive uses that will make overall prosperity increase, eventually leading to increased (not decreased) overall employment. Yes, that's right, you should be welcoming Wal-Mart with open arms if what you want is more jobs (and yes, Krugman surely is aware of this).


Dezakin writes:

Why cant people make criticisms of wall-mart that actually make some economic sense. Perhaps attack wall-mart not on its productivity enhancement which is obviously good, but on the evolution of a natural monopoly that then can just charge economic rent for having the buy in of an economy of scale and quash future competitive threats.

Not that I believe that thats the current situation with Wall-Mart.

Chris Bolts writes:

It's sad that Krugman has traded his impressive economic credentials to become a ignorant fire-breathing liberal elitist. This is nothing but anecdotal evidence, but every Wal-Mart in the Phoenix Metropolitan is either located in areas where businesses thrive off the consumers it brings or businesses will locate where a Wal-Mart will pop up knowing that it will bring consumers. This in turn means new jobs are being created in sectors that are unrelated to retail (even though I have seen several supermarkets and novelty shops locate near Wal-Mart and successfully compete).

It's ironic that not only am I one of the biggest proponents of Wal-Mart, I am one of its biggest detractors because of one simple facet: out of all the Wal-Marts I have visited, only was not overcrowded, cluttered, and a pig sty. For those reasons I limit my business with the Waltons to Sam's Club. At the same time, I don't waste my time holding up signs out in front of a Wal-Mart to "clean up your stores or get out of our neighborhoods" because I recognize the value that Wal-Mart brings to its core constituency: low-income Americans (which I thought is whom people on the left were supposed to champion for).

John S, writes:

Krugman says:

Because Wal-Mart's big-box stores employ fewer workers per dollar of sales than the smaller stores they replace, overall retail employment surely goes down, not up, when Wal-Mart comes to town.

Really? Suppose that before Wal-Mart, the average number of workers per dollar of retail sales is X, then after Wal-Mart it's pX, where p is some percentage greater than 0 and less than 100. Now suppose total retail sales before Wal-Mart were S. My guess is, total retail sales increase when Wal-Mart comes to town. If total retail sales post-Wal-Mart are greater than S/p, then total retail employment will increase, not decrease.

English Professor writes:

The most important thing I've noticed about WalMart is the way it has driven down prices across the board in retailing. If I go to shop in a one of the older department stores (like Marshall Fields or Carson Pirie Scott in Chicago) I am amazed at how prices have come down over the years. You can still find lots of big-ticket items in these stores, but in the kinds of goods where they compete directly with WalMart and Target (e.g., cosmetics for women, or men's and women's casual clothes), they have dropped their prices and are frequently competitive. I suspect that there is a retailing hierarchy effect going on here: Walmart has the lowest prices and the lowest quality goods. Target is a bit higher in both quality and price, but they can't go much higher in price because potential buyers might defect to Walmart. And the department stores have to keep their prices down in order to compete with Target. This strikes me as an economy-wide phenomenon that keeps retail goods accessible to everyone at lower prices than would be the case without Walmart. I would argue that Walmart's success benefits consumers who don't shop there. Cheaper prices overall raise the standard of living of all Americans, but especially of the poor.

William Woodruff writes:

Please read REAL GUY's post. We all have access to this information. Real wage growth is ABSENT in the post 2001 recovery. Is wal-mart solely to blaim, of course not, they are responsibile for only 7% of all retail activity. It is worth noting however the stagnation in real wages.

English Professor writes:

Mr. Woodruff tells us all to look at the post above from the Bureau of Labor Statistics. Well, I'm no economist and no statistician, but I have a couple of questions about those statistics and the supposed decline/stagnation in current wages. To begin with, it always seems problematic to draw conclusions from a single quarter, so when I looked at the statistics in search of a longer time frame, I found only one piece of data: for "nonfinancial corporations," real hourly compensation was -0.6 for the quarter, but it was up 1.5 percent from the same quarter a year ago. Now, that looks like a gain to me.

But here's my real problem. From the way the statistics are presented, it appears that productivity is being treated as if it were a bad thing. If you look at the labor sector titled "business," here is what the statistics appear to be saying: Productivity (for the third quarter) increased at 5.4 percent. This appears to be due to two factors: hours worked declined 0.4 percent, and output increased 5.0 percent. This brings us to the decrease in "real hourly compensation." This decrease is not based on anything like a loss in buying power brought on by inflation; rather, it reflects the relative changes in output and hourly compensation. That is, output increased by 5.0 percent, while "hourly compensation" increased by only 4.2 percent. This results in a fall in "real hourly compensation" of 0.8 (i.e., a reading of -0.8). Well, this strikes me mainly as a statistical artifact. Hourly compensation increased substantially. The only "problem" is that productivity increased even faster. From this I am being told that I should infer economic stagnation and the impoverishment of the working class. From everything I've ever read about economics, increases in productivity are beneficial to the entire society, yet these data are being used as if that concept were no longer valid. That may be the way macroeconomists think, but I am not impressed. And I would like to see how this works out over longer time periods.

Movie Guy writes:

English Professor,

It might help if you studied the long term trends. Considerable public and private data is available.

There is no mystery with regard to what is happening.

I cited the second and third quarter data because we are now on a downward slope for net compensation relative to inflation. Similarly, we have been on a negative path for real wage and salary income for a while.

The near term future outlook appears to be negative based on projected and announced corporate/company changes in health care and pension support in dollar terms. This compensation trend is expected to grow more negative based on the forthcoming impact of the Delphi Effect and similar moves in other industries.

1. Do you believe that workers' real wage and salary income relative to inflation is increasing?

2. Do you believe that more workers as a percentage of the employment workforce are covered by employer-provided health insurance coverage as compared to year 2000?

3. Do you believe that more workers as a percentage of the employment workforce are covered by employer-provided pensions plans of any type, and employers overall are offering higher percentages of support than was the case in the year 2000?

The answer to each question is No.

By the way, your analysis of productivity overlooks the effects of the wedge gap between growth in productivity and real hourly compensation.

Movie Guy writes:

English Professor,

I'll save you a little time.

Here's some basic info, primarily from one source. Yes, the source provides a labor orientation, but the analysis is of survey and tabulated data used by the BLS and other departments of government. It's not likely that some others on this blog will help you see the rest of the story on hourly compensation.

Do note the graphics.

If you find any errors in the following data presentations, please note them on the blog and contact the source.

Employer-Provided Health Care Insurance Coverage:

Fourth consecutive year of decline in employer-provided insurance coverage
October 20, 2005

Employer-Provided Pension Plans:

Google search for current trend info

Kindly check it out. Note Delphi, GM, Ford, Verison as obvious examples of recent trend leaders with regard to where American corporations and companies are likely to go in face of advanced global trade patterns and competitive pressures. The downward support game is on.

National Compensation Survey
BLS, March 2005

Pensions under assault
May 28, 2002

The lack of coverage stems from two related problems: many employers simply do not provide pension plans, and those that do often exclude some of their employees from coverage. While 58% of private sector employees work in a firm that provides pension coverage, only 44% are actually covered. And the lower an employee's income, the less likely he or she is to be covered. Seventy-six percent of workers earning more than $1,000 a week were covered by an employer plan in 1999, but only 33% of workers earning $300-399 a week were covered (U.S. DOL, 2000). Despite $100 billion a year in federal tax subsidies to encourage employers to provide pension plans, the percent of employees covered is no higher now than it was in 1970.

Retirement security - facts at a glace

According to data from the Federal Reserve, the share of households that had a defined contribution plan, such as a 401(k) plan, rose by 70% from 1979 to 1998, while the share of households covered by a traditional defined benefit plan declined simultaneously by 22%.

The growth in defined contribution plans has also made it difficult for lower-income workers to participate in pension plans. In particular, many defined contribution plans require that workers contribute some of their own money to an account, which can be a substantial hurdle for low-income workers. In fact, about 20-30% of people who are eligible to participate in a 401(k) plan don’t contribute.

In 2000, just under 50% of all private sector workers were covered by a pension. However, coverage varied widely depending on a variety of factors. Hours worked, income-level, and education are key factors that influence pension coverage. In 2000, 73% of high earners (the 20% of workers with the highest incomes) were covered, while just 18% of lower-wage workers (the 20% of workers with the lowest incomes) had a pension. Just 14% of part-time workers were covered compared with 51% of full-time workers. Only 25% of workers with less than a high school degree are covered, while 67% of those with an advanced degree are covered.

Rates of pension coverage are higher among union workers. In 2001, 70% of workers under a union contract were covered.

Pension coverage by income quintile, 2000

Employee Wage and Salary Income:


Consumer Prices vs. Unit Labor Costs
December 13, 2005

The productivity and real compensation values come from the nonfarm business productivity accounts of the BLS. The real growth is the wage and salary component is from the Employment Cost Index.

Job quality begins to recover
June 29, 2005

Annual changes in real hourly wages over the past year (from 2003 to 2004)
April 20, 2005

Productivity growth and profits far outpace compensation in current expansion
April 21, 2005


Nominal growth of wages, employment cost index, Wages and Salaries, 1984-2004
October 29, 2004

Separate chart shows nominal wage growth compared to inflation, 2000-2004.

Productivity growth hasn't resulted in increased family income in recent years
September 8, 2004

Worker compensation lagging behind productivity gains
August 18, 2004

Decelerating wages, accelerating benefit costs (2000-2004)
August 4, 2004

Nominal growth of inflation and median weekly earnings of full-time workers, 1994-2003
February 5, 2004

See Figure 1.

Jobs shift from higher-paying to lower-paying industries
January 21, 2004

In 48 of the 50 states, jobs in higher-paying industries have given way to jobs in lower-paying industries since the recession ended in November 2001. Nationwide, industries that are gaining jobs relative to industries that are losing jobs pay 21% less annually. For the 30 states that have lost jobs since the recession purportedly ended, this is the other shoe dropping - not only have jobs been lost, but in 29 of them the losses have been concentrated in higher paying sectors. And for 19 of the 20 states that have seen some small gain in jobs since the end of the recession, the jobs gained have been disproportionately in lower-paying sectors.


Average wages in growing and contracting industries, end of recession through November 2003

Growth in nominal median weekly earnings and inflation, 1999q1-2003q1
April 30, 2003

See second chart.


Earnings Growth Rate by Decade, 1890-2000


Bureau of Labor Statistics

ivan writes:

Does Krugman actually know why Wal-Mart pays lower wages? Is it because profits will be higher? Nope, Wal-Mart's profit margin is razor thin, especially compared with such high wage paying companies like Microsoft. Wal-Mart lower wages leads to lower prices with benefits millions and millions of consumers. Here is Kerry-advisor Jason Furman:

There is little dispute that Wal-Mart’s price reductions have benefited the 120 million American workers employed outside of the retail sector. Plausible estimates of the magnitude of the savings from Wal-Mart are enormous – a total of $263 billion in 2004, or $2,329 per household.2 Even if you grant that Wal-Mart hurts workers in the retail sector – and the evidence for this is far from clear – the magnitude of any potential harm is small in comparison. One study, for example, found that the “Wal-Mart effect” lowered retail wages by $4.7 billion in 2000.
Arthur Solvang writes:

Job loss is cumulative and Wal*Mart does indeed destroy jobs. A typical store offering, say sporting goods, will hire an employee who needs enough money to pay the rent, insure the car, feed oneself and clothe oneself. That store will have employees who usually provide a higher level of knowledge about the items they sell and the price is often higher that an equivalent item found in a Wal*Mart. Note that I stated equivalent item as Wal*Mart usually sells only items on the lower end of the price and quality spectrum. The Coleman camp gear may be a high-profit item for a Sporting Goods store that makes most of its sales on higher quality items.
When Wal*Mart corners the market on cheap sporting goods the specialty store loses a lot of its revenue making it difficult, if not impossible, to remain in business at the same level. Wal*Mart is not substituting one employee for another any more than they are substituting one item for another.
In fact, the items sold are the comodity items and the employee replaced is a comoditity employee. Wal*Mart employees typically do not earn enough money to live with food, shelter, legal vehicles, adequate insurance, etc. and thus put a greater strain on the public resources.
Cumulative job loss includes the manufacturers who make higher quality products that are forced to compete with the cheap products available at Wal*Mart. As those employees are put out of employment there is even greater loss of quality jobs.
In short, Wal*Mart is not competing item for item in efficiency but rather overwhelming the competition with cheap products and cheap labor. One job loss of an adequately paid employee to a cheap employee shovelling comodity items is accuentuated by other job losses of persons who manufacture quality items that cannot be made as efficiently. In addition are the supporting services, restaurants, health-care, housing, insurance and vehicle sales that also suffer when quality income is no longer available.

resigned writes:

I don't know if anyone is reading this, but the problem seems to me that our society is phase separating. For a long time, there was a set of jobs in the manufacturing sector which allowed for people (like my grandparents) with relatively low education levels to obtain good salaries and good benefits. Unfortunately, as these jobs evaporate due to productivity gains, the displaced do not have the skills to transition into the new jobs. For the moment, Walmart is a good force in that it lowers prices, so that its employees have a better chance of being able to afford goods and services. However, it also adds hidden costs to a community in the form of health care burden. However, for the society it is better to accept this burden temporarily (while trying to reform health care) and to encourage retraining. In this transitional period, I think we should acknowledge that it is painful for people and that some may not be able to recover after their jobs are obviated. What worries me is that a number of the good jobs of the future will require higher educational levels, and as there is a distribution of talent in the population, not everyone is going to be capable of those jobs. The best we can do is to assure the best educational possibilities for people as possible and hope that the drop in prices and a social net helps those at the bottom from suffering too much....Step of soapbox...

Comments for this entry have been closed
Return to top