Arnold Kling asks the question:

"If in 1900 you had used previous economic growth to predict growth in the twentieth century, how far wrong would you have been?"

Geez, he's good! I hope the kids he teaches appreciate that! :-) That got me thinking (that's what good teachers will do for you), and led me to prepare the following:

1) A table of the annual per capita GDP growth for the period from 1600-1900 based on DeLong (1998):

Time period...Percent Annual Per Capita GDP Growth

1600-1650.......................0.12

1650-1700.......................0.18

1700-1750.......................0.16

1750-1800.......................0.18

1800-1850.......................0.87

1850-1900.......................1.65

2) Plot those six data points, with the end year on the x axis, and the percentage annual growth on the y axis.

3) Then, draw a straight line regression through the last three data points. The regression line gives an equation of y = 0.0146x - 26.195. If I plug in the year 1950, which is the middle of the 20th century, I get an annual per capita GDP growth rate of 2.28 percent.

So that's what my prediction of the average per capita growth rate for the 20th century would be, based on the data up to the year 1900.

So what was the actual annual per capita GDP growth rate for the 20th century? It was 2.29 percent! In other words, I would have been virtually **dead-on**...accurately predicting per capita growth in the 20th century, based on data up to 1900. Smokin'!

4) What would that regression line do for predicting growth rate in the 21st century? To get that answer, we plug the year 2050 into the equation y = 0.0146x - 26.195. We get an average annual per capita world GDP growth for the 21st century of 3.74 percent.

5) That got me to thinking that I ought to add the data from the 20th century to the 6 data points I had previously. The data points for the 20th century are:

Time period...Percent Annual Per Capita GDP Growth

1900-1950.......................1.76

1950-2000.......................2.83

6) Now, determine a regression line for the five data points ending in years 1800, 1850, 1900, 1950, and 2000. The regression equation is y = 0.124x - 22.033. Plugging in the year 2050, we get a world average annual per capita GDP increase for the 21st century of 3.39 percent.

The whole mess looks like this:

.

.

.

.

.

.

.

7) But even in the face of this information, I'm going to rely on the analysis of Ray Kurzweil. I'm going to expect that computers will be comparable in ability to a human brain in the 2020-2030 time frame. So I'm going to boost my predictions of per capita (P/C) economic growth for the 21st century way beyond those I made in December 2003. My new predictions are:

Time..........Annual P/C GDP growth...Annual P/C GDP growth

Period..........Oct 2004 prediction......Dec. 2003 prediction

2000-2010................3.0...........................2.5

2010-2020................3.5...........................3.0

2020-2030................4.5...........................3.5

2030-2040................6.0...........................4.0

2040-2050................8.0...........................4.5

2050-2060...............11.0...........................6.0

2060-2100........the differences keep growing...!

8) When I plot my new (October 2004) predictions with my old (December 2003) predictions, as well as the extrapolation of the linear regression from the 1800-2000 data...and the predictions of Arnold Kling, Wilfred Beckerman, and Jesse Ausubel...I get this mess:

.

.

.

.

.

.

.

.

Or expressing it in per capita GDP, rather than per capita GDP growth rate, I get this:

Economic growth in the 21st century will be spectacular. Growth will be way beyond anything in the "economic literature," as reported by the IPCC. By the end of the century, average world per-capita GDP will be well over $20,000,000 per year. In fact, world per-capita GDP will cross $100,000 sometime around 2050. This has huge implications for just about every aspect of thinking about the 21st century. More on the implications later. (I know I promised to discuss implications at the end of my "second thoughts"...but I really mean it this time.)

## Recent Comments