This is the second part of a series of posts responding to Matt Steinglass at the Economist, who put forward some common progressive arguments in favor of more unionization recently. To understand the big picture problem with the argument for more unions, it’s useful to look at the reasons why unions have died out in the first place. One common explanation is that the most highly unionized industries, like manufacturing, have shrunk, and that unionization was simply taken along for the ride. This is not the case however. As Barry Hirsch details, while overall manufacturing employment fell from 20.1 million from 1973 to 15.6 million in 2006, nonunion manufacturing employment rose by 1.5 million. A similar patter persisted in the other main union industries of construction, and transportation/communication/utilities: union employment fell while nonunion employment rose. The graphs below, from Hirsch, tell the story pretty clearly:
If declining industries is not the cause then what? Hirsh identifies 3 main explanations for the decline of unionism: competitive, structural, and institutional. Ultimately, he provides a convincing argument that the fall of unionism is due to a more competitive and dynamic economy. Part of the problem, he argues, is that collective bargaining slows firms down:
“Were changes in the economic environment very gradual and competitive pressures weak, a formal and highly deliberate union governance structure might pose few problems. The costs of deliberate or sluggish union governance, however, increase with the speed of change and the degree of competition. New information is constantly coming to a firm and its workers and it is prohibitively costly. to have explicit contract terms for every possible contingency. Revising formal contractual terms is costly. Although many collective bargaining agreements have broad management rights clauses, formalized contractual governance limits flexibility and managerial discretion in union companies. “
A related problem is that by cartelizing labor, unions raise wages above the competitive level. The more competitive a firm’s market, the less they will be able to raise price in response to higher wages, and the more costly unionization is in terms of lost employment. For both reasons it is difficult more difficult for unionization to exist in a competitive and dynamic economy.
A common counterargument to this, which Matt makes, is to point to the high employment and unionization of the 50s and 60s as evidence that they can co-exist. But the economy was much less dynamic and competitive than it was today. As detailed by Brink Lindsay in Nostalgianomics, labor market competition in that time period was limited by discrimination against minorities and women, and strict immigration restrictions. Goods market competition was limited by government policy like tariffs and price controls, and a much lower level of global competition than you have today. Despite this, strong economic growth was possible due to lots of long-hanging productivity fruit and innovation.
Today low unemployment can’t co-exist with high unionization of the kind we have in the U.S. because the world, and especially the U.S., is more competitive and dynamic. I think an implicit argument made by some, although not necessarily Matt, is that we should be willing to give up the dynamic and competitive economy in exchange for more unionization. But aside from being good for all the reasons detailed in Nostalgianomics (free trade, less discrimination, and more immigration are good things), a competitive and dynamic economy is extremely important. As Tyler Cowen argues in The Great Stagnation, we’re out of low-hanging fruit like we used to have, and so we can’t afford to give up the economic growth we do have by attempting to decrease labor market and goods market competition enough that unions can thrive again.
10 comments
Comments feed for this article
Friday ~ May 20th, 2011 at 11:51 am
Brennan
Canada has a much higher (although still declining) unionization rate than the US, and operates in a similarly advanced economy. Its unions also operate under similar rules to the US (unlike the UK or Germany, for instance, which have very different legal structures), although it is easier to form and maintain a union there. Going beyond Canada, how does this take into account the much higher rates of unionization in Germany and Europe more generally?
Any explanation that doesn’t deal with these issues seems to be missing something important.
Friday ~ May 20th, 2011 at 12:36 pm
BSE
Your last post was better. Before I get into anything else, let me point out that a decline in labor employment is evidence for nothing other than a decline in labor employment. For example, in Germany, propensity for a firm to export is correlated with unionization. How does that square with you argument that union firms are “uncompetitive”? Now what I am going to say applies to the anti-union crowd, which you say doesn’t include you (you are just making a counterargument to Mr Steinglass).
I’m with Krugman on this whole “competitive” meme. Even if international “competition” is a real constraint on national economies, the effect is marginal at best. Even in this globalized world of ours, employment in exporting industries is miniscule, and the volume of trade is much smaller than theory predicts that it “should” be.
You also keep referring to the economy as more “dynamic” and “competitive” than it used to be. There is in fact little evidence that either is actually true and some evidence that both may be false. TFP growth slowed after about 1970; about the same time that de-unionization started. Not to mention that construction saw a similar trend as manufacturing, but is not subject to the same kind of competition.
Back to the TFP point: for any economist this should be cause for reflection – one thing we agree on is that unions improve productivity (and when it comes to unions economists no longer argee on much)! Instead, you invoke the “low-hanging fruit” argument. If the evidence for a more “dynamic” and “competitive” US economy are weak, the evidence for the “low-hanging fruit” is all but non-existent. As far as I can tell the ENTIRE sum of evidence for “low-hanging fruit” is in the TFP result I quoted. Nonsense. Absolute nonsense. If the “low-hanging fruit” hypothesis were true, then we would see a slow transition to the lower growth rate from the higher. But that’s not what we see. Instead, we see a “elbow” in the TFP data. That’s not EVEN getting into why would this happen in the ’70s: just before the computer revolution. In the name of Zeus’s too-tight pants, why would computers not boost the TFP trend!?! Anyway, I digress.
To the competitive issue (between firms, not international… FYI in this post you refer to both interchangeably which is confusing); it is true that a unionized firm is at a disadvantage compared to a non-unionized one all else being equal, but what you miss is that this used to work both ways. It used to be that nonunion firms would be under pressure to unionize or raise wages to compensate. So why did THAT change? My view here is that the power of unions was broken long before they actually disappeared. There WAS a major change in US labor policy around 1970 (’73?… anyway, its not important exactly when it happened) making it harder to unionize. The result, competition only went one way and unionized firms started dying. This explains the trend, and it also explains the trend in the cross-country data. Europe is losing unions too, albeit from a higher level than we are. For them, though, it didn’t start in the ’70s like here, but in the ’90s. Coinciding perfectly with neoliberal reforms to the labor market.
Now, this can be a good or bad thing and unions are complicated. Only the free-market ideologues would assume that a “freer market” must mean a more efficient market. It may happen (and I would argue that it does happen) that a “cartel of workers” compensates for a too-low market wage. The problem is pricing power; in manufacturing especially, workers have little and management a lot. I won’t get into models (or data) of why here, I’ll only ask this: what other options do less educated people have?
Friday ~ May 20th, 2011 at 2:11 pm
TomB
I am anti-public sector union, slightly anti-private sector union.
I agree with BSE on a number of things: your last post was better, decline in unionization doesn’t tell us much, the “low-hanging fruit” argument is not very persuasive and not really supported by the evidence, and unions can be good or bad.
I think that if unions are formed at all the companies in an industry, the unions more often that not end up distorting the incentive structure and harming growth. Look at the automotive industry. After lower tariffs and cheaper shipping came about, international competition in the automotive industry is real. The unions took a hit because they locked inefficiencies into that industry in the U.S.
That said, if some only some firms are unionized, I think those firms could result in increased growth or productivity. The higher wages would attract the better workers in that industry. The union could help smooth over management and labor relations, and all the other benefits unions can sometime offer. The union is also disciplined by the market in the sense that if they negotiate too much, non-union firms can take over their market share.
If only some of the firms in an industry are unionized, the negative effects of an above market clearing price are reduced, because some firms are offering a lower price (wage).
I think too many variables are at work to attribute the decline in unionism to any inherent problems with unions. However, if the benefits provided by unions really outweighed the costs, I think the decline would less. My opinion (not based on data) is that the real issue is that it is more likely that a union will become a drag on a company than that the union will be effective run so that it can provide benefits.
Friday ~ May 20th, 2011 at 2:28 pm
Matt Waters
I think the issue with American unions is how they are set up to be more adversarial than unions in other countries. I know Japan has separate unions for each car company, while the US just has the UAW. In the former case, the union is much more intertwined with the fate of its corporation. In the latter case, the union does not have an incentive to work as much with the corporations to be competitive. The UAW would rather try to have a monopoly on labor than for their local chapters to be more competitive.
Could we move to a more German model for labor? Maybe, considering that the UAW has become far more welcoming to the Big Three’s demands once they realized they really could go under. But the old-school, 1950′s UAW people long for just cannot coexist with a competitive economy. If private-sector unionization ever comes back, it will need to get past the old mentality of a labor cartel.
Friday ~ May 20th, 2011 at 9:05 pm
Jim Glass
“A common counterargument to this, which Matt makes, is to point to the high employment and unionization of the 50s and 60s as evidence that they can co-exist. But the economy was much less dynamic and competitive than it was today.”
Yes — but I think people also underestimate simple falling demand for union services on the part of workers.
In the 50s and 60s industrial union members were usually grade-school or maybe high-school educated in the 1920s or 30s, living in a world where they needed some oraganization to create pension for them, a credit union for them to save and borrow at, and so on.
Today even production line workers are educated in statistical analyis, can set up their own retirement accounts, have a world of investing opportunities open to them via their online brokers and so on.
Why would they want to pay a percentage of their wages to a union to do for them what they can do better for themselves? (Especially when they know it will make their employer less competitive.)
That is, workers are buying less in the way of union services for the same reason they buy less of other things — they don’t want them.
Saturday ~ May 21st, 2011 at 7:11 am
TheStone
Jim G. – I spent several years working on and near production lines and never discussed statistical analysis or swapped notes on online brokers. And most of the newer hires on the line do NOT have retirement accounts of any sort. In fact, many of them are not employees of the company that owns the production line, but rather of a temp agency. If a more dynamic and competitive economy means everybody working for a temp agency, or under circrumstances resembling working for a temp agency, I don’t want any part of it. Nor do I believe that such a development was entirely unavoidable. The post that we are all responding to assumes that these conditions fell from the sky and was not associated with conscious decision-making by any human beings or institutions and I just don’t think that’s true. Other commenters have touched on that in more depth, so I’ll leave off here.
Friday ~ May 20th, 2011 at 9:25 pm
Koheleth
Classic correlation-causation problem. If by “dynamic” you mean that production is more mobile, then we have a very different possible explanation for the decline in industrial unionization: the rise of “right to work” states has promoted a race to the bottom as firms hopscotch geographically, shopping for the most favorable terms. Meanwhile “just-in-time” supply chains facilitated by advances in communications mean that parts can suddenly be sourced from a wide variety of locations. Purely conjectural, but it is supported just as well by the data as your arguments about natural selection for non-union labor.
Saturday ~ May 21st, 2011 at 12:10 am
Doug Weinfield
To totally ignore the very significant efforts made by many large corporations, and organizations, such as the Chamber of Commerce, to bust unions, is so staggering as to make the rest of the posting useless.
Saturday ~ May 21st, 2011 at 8:09 am
Browsing Catharsis – 05.21.11 « Increasing Marginal Utility
[...] Overall manufacturing unemployment is down in America since 1973, but nonunion employment is actuall…. [...]
Thursday ~ June 9th, 2011 at 12:55 pm
The Economics of Unions - E.D. Kain - American Times - Forbes
[...] Ozemik has two thoughtful pieces up (here and here) on the economics of unions and the difficulty organized labor faces in a global, dynamic economy. [...]