it does, but not well enough to render “real GDP” reliable as a welfare measure. for example rents captured as reductions in quality, except within the particular items and ways hedonics can quantify them, don’t pass through to price. 1/

horrible customer service, for example, doesn’t hurt GDP, assuming the demand-side is unaffected by reduced call center employment, which is a good assumption when macro policy targets demand. /fin

use of gdp as welfare measure was always based on qualitative correlations. gdp’s (well gnp’s) inventor said it should not be so used, but seemed to work pretty well. 1/

but that is changing, because of increasing market power in especially the US economy, and relatedly the difference between cost-based and market-based pricing in GDP for govt purchased vs privately purchased goods and services. 2/

under consolidated markets, GDP comes to include rents captured by monopolists. rent extraction reduces welfare but scores as higher GDP. 3/

relatedly, the cost paid for government purchased healthcare in social democracies is much, much lower than the “market prices” of healthcare in the US. quality and outcomes are not broadly higher. 4/

so, US health pathology increases US GDP, while failing to represent an increase (arguably representing a major loss) of real welfare in the United States. /fin

one take: if the government offers a higher wage but still lower than marginal productivity, a profit-maximizing (former) monopsonist raises its wage. the worker then is only actually lured away if the government offers a wage above the true opportunity cost, presumably because the work is worth it.

(in a cost-benefit analysis, analysts unaware of the labor monopsony might erroneously judge a program worthwhile by underestimating the costs that would actually be required to hire the labor.)