Per capita GDP increase over time (when controlled for quality) is around 36 times in 100 years

GDP serves an excellent purpose, and its annual growth is a useful indicator of improved wellbeing and prosperity.

However, comparing GDP cross-sectionally across countries is hard. The exchange rate applies only to the traded sector, so comparisons using this value are fraught. So the purchasing power parity (PPP) methodology corrects for this difficulty. I generally prefer PPP calculations to exchange rate-based comparisons of GDP.

There is another major difficulty, however. While technology can be assumed to be similar over two years, it is actually changing quite rapidly. Comparing GDP growth across time (time-series) within the same country is fraught, even if we control for inflation using the consumer price index (using CPI is fraught, as well).

The issue of quality becomes quite serious when we compare the market value of output across more than 2-3 years.

Let me illustrate.

My father purchased his first car (a Fiat) in 1970. The interior of the car was tolerably comparable to the interior of modern cars. However, its engine (of 1.1 litre, I think) and mechanics was very poor. The car broke down frequently.

35 years later I purchased a Toyota Avalon (3.5 litres engine). This car has never broken down for the past ten years. I perform the usual servicing (around once in 15-18 months) but it is sturdy like a rock. To get such a car in 1970, my father would have had to pay a lot more (even assuming that such quality was available in the market).

The same applies to telephones/mobile phones, washing machines, computers: things that have significantly improved in quality as a result of technological change. 

Question: How can one compare quality changes over time in an meaningful way?

if a buyer in 1970 had a choice between a 1970 Fiat and a 2005 Toyota Avalon, what would such a buyer have paid for the Avalon? An alternative way is: How much would someone in 2005 pay for a 1970 Fiat? This is far more practical.

I paid $19,000 for my 1.5 year old Toyota Avalon. I would not have paid more than $500 for the 1970 Fiat. Its value in 2005 would be little more than scrap metal, even as a brand new car.

If one controls for engine size, etc. one could reasonably conclude that quality had increased by at least 20 times, possibly in the range of 50 times, over 35 years.

Using such market-based approaches, we could compare high value goods across time, thereby attributing a quality factor into the GDP calculations. 

Note that there has been very little change in teh quality of a number of things: food, bricks, houses. So the overall quality inflator for GDP would be much lower than 20, possibly in the range of 3.

But even that (3) is HUGE. In merely 30 years, the true TIME-CONTROLLED MARKET VALUE of what I own has increased by 3 times, just because of an improvement in technology.

Over 100 years, per capita income, as measured by GDP, generally increases by around 6 times. But if my calculation of quality is approximately correct (note that quality is accelerating rapidly), that adds a quality inflator of around 6 times over 100 years (3 x 2 x 1).

This leads to a net improvement in REAL prosperity of around 36 times over 100 years.

This is merely the sketch of a possible methodology to assign a market-based measure to quality improvements over time. I’m sure others have explored this topic in much greater detail. Do let me know about any such work you know about.