Schumpeter's Calculation and the Present State of Things

In his great book Capitalism, Socialism and Democracy, Joseph Schumpeter envisioned further economic growth eliminating "want" (an unfortunate choice of word, since wants are really limitless, but what is meant is destitution or something like that) in just a few decades. The kind of economic growth which would achieve this immense feat would have to be in line with the growth numbers which had been seen during the past century or so up to his writing the 1947 edition of the aforementioned book. On page 67, for instance, Schumpeter notes the availability of modern dentistry to workmen which Louis XIV would not have obtained even if he had spent all his wealth. On page 69, Schumpeter continues:

"Now if the system had another run such as it had in the sixty years preceding 1928 and really reached the $1,300 per head of population, it is easy to see that all of the desiderata that have so far been espoused by any social reformers - practically without exception, including even the greater part of the cranks - either would be fulfilled automatically or could be fulfilled without significant interference with the capitalist process. Ample provision for the unemployed in particular would then be not only a tolerable but a light burden."

The emphasis is in the original. According to the BLS's inflation calculator, $1,300 in 1928 is $17,963.80 in 2014, well below even half of the American GDP per capita. So by the Great Austrian's reckoning, spending on transfer payments should be a very low fraction of GDP, indeed. Yet the welfare state consumes a lot more today than it ever has before. For instance, Charles Murray's Coming Apart cites the figure $1,500,000,000,000 (called a trillion and a half in America; I wonder if numbers have ever got this big in the UK, but Britons would call the figure one and a half billion) spent purely on income transfers in America, and throughout the West, total government spending as a share of GDP is typically around 40 per cent.

 

Now consider the $1.5 billion (or "trillion") figure and relate it to what Schumpeter had to say about eliminating want. It comes to almost $5,000 per American (of whom there are almost 320 million according to the Census Bureau), which is not that much less than one third of the $1,300 per-capita GDP in 1928 that Schumpeter nicely argued would suffice for wants to be eliminated "without significant interference with the capitalist process". But almost a one-third of per-capita GDP being spent on transfers is surely significant interference. Yet there remains poverty, although it is incredibly rare by world standards. What gives?

 

These facts indicate to me that the welfare state is not really about helping poor people. Moreover, there is the suspicion that there are laws in social science which make it incredibly difficult for welfare programmes to fulfil their ostensible purpose. One such candidate is Director's Law, named after University of Chicago Law Professor and Economist Aaron Director. It is not really a "law" as much as a vast and coherent set of empirical findings, for instance that tuition at fancy universities is often subsidized even though college students typically come from fairly affluent families.

 

However, there is some logic to Director's "Law", independent of observation. If a welfare programme is designed to help the poorest decile (say), it is plain that this is paid for mostly by persons far richer that that. The poorest decile is not a very useful group for office-seeking politicians, who must seek the approval of many more people than these in order to gain power (besides, people this poor tend not to vote much). Consequently, the winning proposals will have money shifted from fewer people than 90 per cent of the population, and to more people than just the poorest decile. Programmes specifically directed at the poor will tend to go unfunded.

 

What about expanded programmes? Everyone wants to be a beneficiary of such programmes, so they might easily grow a great deal, but once they have grown enough, the bottom ten per cent may no longer be such a valuable part of whatever coalitions form. Expanded programmes are palpably more attractive to interest groups than are narrower ones, and so those who can organize politically will tend to secure the benefits for themselves. Thus, farmers and other small and non-poor groups appropriate a lot of transfer money by good lobbying. Again, the poor folks lose out.

In addition, the poor may tend to be disadvantaged in utilizing whatever programmes for which they qualify. To get good health care, it helps to be articulate and to already know a bit about the system (for instance, by having friends within it), traits which the poor may well possess, but perhaps not as much as middle-class individuals. Services directed at poor areas may also be at a disadvantage when it comes to attracting talent to provide them. I can imagine good teachers shunning schools in poor areas. Raising the salaries in those schools will attract better talent, but as has been shown, this is politically difficult to do.

Schumpeter's calculations were surely correct and, on merely technical grounds, poverty could be eliminated today even with a much smaller government. It just happens that that is not how politics works.