Thursday, June 13, 2019

Baumol Effect Recognition is a Good Starting Point

A recent back and forth regarding the Baumol effect among various bloggers (and their commenters) is quite encouraging. For instance, Arnold Kling (here, here and here) and Bryan Caplan also responded to Alex Tabarrok's posts, Scott Alexander reviewed the book, and Scott Sumner adds some thoughts re healthcare. Just the fact an important structural feature such as this is getting some attention - when distraction tends to be the norm in times of political and cultural turmoil - is cause for celebration.

For that matter, services which contain substantial time centered components, haven't really benefited from serious consideration. Yet this is somewhat understandable, since service sector dominance only came to the fore (at least in recent historical memory) in the latter part of the 20th century. For the most part, only a decade earlier, it seemed time centered services were little more than an afterthought in most online discussions. I recalled some discussion for instance about low skill service providers such as restaurant and retail work. In recent years the dialogue is turning towards high skills services capacity. And time centered knowledge or high skill services as final product, may include economic ramifications we've only begun to discern.

Fortunately it's not necessary to solve the major mysteries of the Baumol effect in short order! For that matter, it appears to play a partial role in a larger foundation of how aggregate time value has come to be conceptualized. When William Baumol explored this phenomenon decades earlier, he may have been more interested in how it affected wage relationships at local levels, in what could be considered an income smoothing effect for skills coordination among divergent groups in cities or regions. Interestingly enough, he didn't seem concerned it would later prove problematic for the nature of general equilibrium monetary flows.

Part of the mystery is how service sector activity tends to mostly level off at around 80 percent of GDP before stabilizing in a mature economy. Why such a high portion of GDP, if one considers the extent to which (for instance) manufacturing levels also impact service sector levels? Perhaps the wealth of housing assets is a factor, especially since much of this is reflected in the financial sector. How to think about the supply side limits of housing wealth, in terms of productive agglomeration? Might the high skill wage limits of the Baumol effect also contribute to the relative scarcity of productive agglomeration?

The relatively recent dominance of service sector activity is an apt reminder, how young economics as a science really is. In some respects a new process of theoretical adjustment could just be getting started, should we be fortunate enough in the years ahead to maintain today's degree of relative economic normalcy! Meanwhile, the wealth of recent high skill services dominance is a major contributing factor, to the increasingly blurred line between public and private activity. It's incredible to realize how different things were, only a century earlier. By way of example, Paul Samuelson, in Economics, noted the earlier clarity of government roles:
Prior to World War I, local government was by far the most important of the three. The federal government did little more than pay for national defense, meet pensions and interest on past wars, finance a few public works, and pay salaries of judges, congressmen, and other government officials. Most of its tax collection came from liquor and tobacco excises and tariff duties levied on imports. Life was simple. Local governments performed most functions and depended primarily on property taxes for their finance.
It's been a long time since life felt that simple. Still, something interesting about Why are the Prices so D*mn High? was the authors' confidence that the Baumol effect was more of a positive than a negative. Will this prove to be the case? It depends. Can society come to terms with the fact real wage gains will need a different approach in the foreseeable future, given the recent limits suggested by this phenomenon? Hopefully, yes.

No comments:

Post a Comment