When public sector unions participate in the political process, candidates who are supported by unions can become officeholders charged with negotiating collective agreements for public sector unions. The unions, then, have agents on both sides of the bargaining table when their salaries and benefits are negotiated. This obviously creates a situation in which any given bargaining session may yield compensation for union members that is greater than that which would have been negotiated under adversarial circumstances. Terry Moe makes the point more thoroughly here.Stephen Bainbridge has an extensive post on the subject, also making reference to Moe's argument, among others.
The logic of anti-union arguments is pretty clear, in any event. Either public employee unions extract greater wages and benefits for their members than they would have otherwise received (increasing spending) or they do not (no spending effects). In the first case, union busting may decrease public spending. In the second, union busting doesn't hurt anything because the unions aren't winning any rewards for their members.
Yet, John's original analysis suggests there is little link between state spending an unionization. John has since updated his analysis on the basis of readers' comments. John's updates deal with the distinction between public sector unions and unionization overall, deficits and spending, and deficits and debt (relative to the size of state economies). The updates show that the link between deficits and public sector unionization is stronger than the link between deficits and unionization in general, and also that there are signifcant associations between public unionization and spending and between public unionization and state debt.
Having had a chance to update my own data from yesterday (including filling some missing observations) using the sites John links to--- http://www.unionstats.com/ and http://usgovernmentspending.com/---I have a couple further notes. First, I think the association between public sector unionization and spending is a bit strong than John makes out. John plots the percent of public sector employees in unions against state spending per capita to judge the relationship between unionization and spending. I think the issue comes into a bit sharper focus by considering the proportion of all workers in a state covered by a public sector union collective bargaining agreement against spending per member of the workforce. The results, illustrated below, are pretty similar to Johns. I think you get more leverage, though, because this unionization measure takes into account the size of the unionized public workforce relative to the size of the state's entire working population (and not just its public sector). Spending per worker rather than per capita probably makes even less difference than the first, it seems to me to match up with the political claims that tax payers are supporting overly generous compensation packages for public sector employees a bit better than a per capita figure.
What about deficits, though?
To start to get a handle on deficits, I look at state revenues per worker and public unionization (using the same measure as above). The top figure is based on all 50 states plus DC; the bottom figure excludes AK, DC, and WY.
In both cases, we see a positive association between unionization and revenues collected. In the all states case, the correlation between public sector unionization and revenues per worker is 0.40 (p<0.01). Removing the outliers, the correlation is 0.54 (p<0.01). As we might expect, higher public unionization is associated with higher revenues (taxes and fees) as well as higher spending.
However, the association between public unionization and spending is "steeper" than the association between public unionization and revenues. An increase of 10 percentage in public unionization (all cases) predicts $13 in addition spending per worker, but only $8.6 thousand in additional revenues.
How might we account for this disparity?
An obvious answer is that public sector unions exert asymmetrical political pressure on a states; fiscal decisionmaking. They place demands for higher spending, but they do not make similar demands for revenue collections. The second should be implicit in the first, but in practice, revenue policies are (sadly) made independently of spending policies. Irresponsible or inattentive officeholders might easily submit to unions' political demands for higher compensation (spending) and other constituents' demands for stable or lower taxes and fees (revenues), winning accolades form both in the short-run while burying their states in fiscal disasters in the long-run.
My take away, public sector unionization is probably related to higher spending and higher revenue collections. It is probably also related to higher deficits, but only because of other structural features of the political process related to the elector incentives of legislators and the severability of spending and revenue questions. Union busting may reduce government spending, but that's only half the problem. Figuring out a way to induce actual fiscal responsibility in state governments---in the sense of creating adequate, sustainable, and fair revenue streams that support a given basket of spending policies---is the harder part.