Virginia Postrel argues

“Maybe the average hotel guest wouldn’t care whether there are 300-count or 200-count sheets. Maybe they just want a bed that’s not caved in,” says Dan Ginsburg, a supervisory economist at the Bureau of Labor Statistics.

If the guests do not see the difference as more valuable, then the higher charge should be factored into inflation measures. If the redesigned room costs $15 a night more, the bureau’s statistics should count the whole $15 toward an increase in the index.

If, on the other hand, guests care as much about aesthetics as hoteliers believe they do, it would be irresponsible to treat the $15 as a true price increase.

The index problem is gnarly. Much more complicated than most people realize. For one thing it attempts to measure something that theoretically doesn’t exist compiled from market transactions we cannot observe.

For the more technically inclined this is an attempt to aggregate over the expenditure functions of multiple folks. Problem is interpersonal utility cannot meaningfully be aggregated and expenditure functions and their market analog, Hicksian Demand, cannot be observed.

That having been said, back when I thought about this a lot, numerous squeezes seemed promising. That is, there is no way you can get the “right” amount of inflation.

However, you can calculate an amount of inflation that is definitely too high. And, you can calculate an amount of inflation that is definitely too low. Then you refine those estimates until you can squeeze the range into something that is less than measurement error.

So in the end you say, if our measurements are correct (which they are not) then inflation is somewhere between 2.06 per year and 2.09 per year. At that point you have practically speaking, beaten the problem

As it relates to hotel rooms – and I working from memory – the question isn’t simply in “how do customers feel about improved rooms.” It is do improved rooms sell at a premium to non-improved rooms. If they do not then at the margin we can conclude the customer does not prefer them.

If they do, then this places an upper bound on the quality value of the improvement. Something can’t be worth more to the marginal consumer than the market value.

So carefully applying this steps we ought to be able to get some sense of what 300 thread count sheets were worth.

As a side note, if hotel managers chose to go out in the market and by 300 count thread sheets and then could not charge a premium for rooms with them then that would count as a negative shock to Total Factor Productivity.

The manager just decided to do something that destroys value in the economy. To integrate it into a macro-models, it creates inflation by ever-so-slightly pulling back on aggregate supply.