From Comments: The Economics of Declining Union Membership

From comments by Donald Coffin in response to this post on whether the decline in unionization can be attributed to changes in political ideology:

One of the aspects to teaching collective bargaining (as an economist, rather than teaching it as a course in how to negotiate, or how to deal with grievances) is than one has to deal explicitly with the issue of the decline in ... union membership as a percentage of employment... The fact is that union density in the private sector in the US has declined relatively linearly since the early 1960s--from around 35% of the private sector in the early 1960s to around 8% now.

I think there are a couple of explanations for this, and being an economist, it's no surprise that I see these factors in demand-and-supply terms.

On the demand side.

(1) A lot of government policies, from anti-discrimination law through occupational health and safety law (and others--the Family and Medical LEave Act, the Americans With Disabilities Act), have put governments "in the place of unions." These are things that unions used to see themselves as responsible for, but are now government responsibilities. I regard the legislative extension of these protections as a good thing, despite the side effect of making unions less valuable.

(2) In manufacturing in particular, globalization has reduced the bargaining power of US unions. Higher wages and benefits gained today will lead to larger employment declines than they did back in the 1950s and 1960s, when trade was a substantially smaller part of our economy. (Related to this, the percentage wage differential between unionized workers and workers not represented by unions has declined from about 25% in 1984 to about 18% in 2004.) If unions are less able to provide higher levels of compensation, then the demand for union services is likely to decline as well. (It's worth noting that the decline in the union wage differential has declined much more in mining and manufacturing--now about 5% and 7% respectively, than in construction--over 50%.)

In some industries, in which tremendous growth has occurred in total employment, the nature of the labor force also tends to reduce the perceived benefits of union membership. In industries like the fast-food business, the overwhelmingly part-time and young (and therefore temporary) nature of the labor force serves as a prime example, but the increasingly part-time nature of employment in retail is also important.

Between these three factors, it seems likely that the demand for union services is less than it used to be.

Beyond that, though, there is

The supply side.

(1) The change in the industrial structure of the US has not been beneficial to unions. Organizing has historically been easier when unions can work with relatively large numbers of workers in individual potential bargaining units (manufacturing), where there is a strong skill component to work (construction), or where there's substantial geographical isolation (mining). The shift away from goods production (in employment, not in output shares) to services has changed this. Service (retail, business services, personal services) establishments tend to be smaller. It becomes more difficult to organize on a company-wide basis (Wal-Mart, for example), because it's not clear that the current law would allow for identifying Wal-Mart, rather than the individual Wal-Mart store, as the relevant bargaining unit).

(2) The change in the regulatory environment has also mattered. The NLRB is no longer even neutral; it's overtly hostile to unions. This leads to much higher costs of organizing, of bargaining, and of enforcing agreements. This has made it easier

(3) for employers to be more active in their opposition to unions. More employer opposition has also increased the costs of providing union services.

(4) Finally, unions have been relatively slow to adopt cost-reducing (and perhaps significantly labor-reducing) technologies, for what might be termed ideological reasons. It's harder to look like a defender of labor if you're laying off your own workers because you've adopted labor-saving technology.

Add all this up, and the costs of providing union services has increased, which we (as economists) would generally agree will cause the supply curve of union services to shift to the left.

Note that two of these factors do relate to changes in government and business behavior, and could be changed, at least in part, through appropriate legislation.

But given the decline in the demand for union services and the decrease in the supply of union services, the outcome is pretty much what we'd predict.

For my own part, I think the demand side of this is much harder to address than is the supply side. The supply side is partly amenable to legislative action. The demand side is not, really, short of enacting trade restrictions.