Monday, February 19, 2018

Services Go Up, Stuff Goes Down (Except Textbooks)


This chart makes quite the inkblot test.

It shows rates of price increase, adjusted for inflation, for a set of goods and services in the United States over the last twenty years.  Among the costs that dropped: cars, furniture, clothes, software, toys, and tv’s.  Wages, food, and housing were basically flat (though housing is spikier across locations, I’m guessing).  Costs that rose included medical care, childcare, college tuition, hospital stays, and textbooks.

With the exception of textbooks, an easy way to summarize the chart is that stuff got cheaper, and services got more expensive.  It’s Baumol’s cost disease in action.

Notably, the commentary on the economics site that posted it missed that point completely, instead veering into some snark about socialism.  I’m guessing that’s why the chart didn’t include international comparisons; we’d notice quickly that jackbooted socialist dystopias like Canada and Norway have cheaper healthcare and education than we do. But never mind that.  Instead, the site posits a vague conspiracy by which “bread and circuses” distract the masses.  Sigh.

Textbooks bucked the trend, and although they aren’t on the chart, I’d bet that prescription drugs did, too.  They’re both basically unregulated for-profit monopolies, and they behave accordingly.  The little squiggle at the end of the chart for textbooks may represent some overdue and very welcome competitive pressure from OER; I’m hoping to see a lot more of that.  Prescription drugs are another blog post entirely.

The reason that costs of services go up while costs of things go down is the relative difficulty of productivity increases.  It’s easier to increase the number of tv’s produced per hour than the amount a student learns per hour.  When both of those enterprises are in the same economy, the cost of the former will drop more than average, and the cost of the latter will rise more than average.  To the uninformed voter, or economist, or blogger, that will look like superior management in the former sector, and a lack of discipline in the latter.  It’s neither.  

I knew someone in college who had perfect pitch.  Not perfect relative pitch, but the real thing.  I commented that it must have been nice.  No, she said; it was awful.  She could hear every little flaw in everything.  Life was full of fingernails-on-chalkboard sounds for her.  

If you know about Baumol’s cost disease, that’s what a lot of our political and economic arguments feel like.  They’re so far off base that they’re actually painful to hear.  

Why is the public sector chronically squeezed?  Because it consists mostly of services, most of which are time-bound.  Education is measured in years (K-12) or hours (higher ed).  Measuring in units of time defeats productivity increases, by definition.  Incarceration is measured in time, too.  Police, fire, and military protection are 24 hours.  Closer to home, my daughter is in the 8th grade.  The 8th grade takes as long now as it did when I was in it, heaven help us all.  Meanwhile, the opportunity cost of what that teacher could have produced in manufacturing has gone up exponentially.  That’s not the fault of the school board, the teachers’ union, or school administrators.  It’s the nature of the enterprise.

Look at the graph.  Healthcare, education, and childcare (especially if you count the K-12 system in that category) draw heavily on public funds.  Furniture and television production don’t.  Public funds aren’t the critical variable -- again, note the lower cost of healthcare in single-payer systems -- but voters often see a correlation.  They see themselves getting squeezed on necessities, and blame the people who provide the necessities.  It’s understandable, if false.

In the regular calls for increased civic engagement in higher ed, I almost never see calls for explaining Baumol’s cost disease before anyone goes and votes without knowing about it.  We should.  If we don’t, we’ll keep getting blamed, punished, and cut.  The graph speaks for itself, whether its producers know it or not.