Saturday, December 12, 2009

Part Four of Three


You have to go where the numbers take you, that's what I think. But after I finished the first three posts of my "On the Growth of Government" series, I realized that I could have achieved the same result by making just one adjustment to Federal spending. Instead of three.

I adjusted for inflation, population, and living standards. But in the living standards adjustment, I had to disallow inflation and population because those adjustment were already made. So I ended up using "real GDP per capita."

If I didn't adjust for inflation separately, I could use "nominal GDP per capita." And if I didn't adjust for population separately, I could drop the "per capita." So now I want to look at Federal spending growth again, adjusted only for the growth of "nominal GDP."

The accuracy will be much better, I think, using this approach. Using only one tweak.

I really don't see how this "unified" graph -- annual change in Federal spending growth, less annual change in Nominal GDP -- can look like my final graph from Part 3. Because I've heard too many horror stories of "Federal spending relative to GDP." Those stories always tell the tale of excessive Federal spending.

Okay. So now it's even more important to do the new graph.


You know, you can just type DOCS as the URL in Firefox, and it brings up your Google Docs. These guys are good.


For symmetry or whatever, here's the long view:



As expected, this view shows little detail. Just like the first graph in Part One of the series. Following the method of that post, I cropped the graph to remove the WWI and WWII spikes.

As expected, the cropped graph shows a little more detail:



In Part One I noted that most of the numbers fall somewhere between zero and twenty. On this new graph, I'd say they mostly fall between ten and minus ten. (The numbers are lower here, because this graph already has all the adjustments subtracted out.) But as in the earlier post, at this "zoom" level, the detail is still inadequate. So as before, I cropped out the biggest remaining spikes:



This graph shows the annual growth rate of Federal spending, with the annual growth rate of GDP subtracted out. This is the consolidated version of the picture developed in the first three posts of the series.

The jittery blue line shows year-to-year changes, as in earlier posts. The red line shows the five-year running average. This graph is very similar to the final graph of Part 3. Spending growth wanders back and forth across the zero-line.

I am relieved to see this similarity. It means I didn't accumulate a lot of "rounding errors" in the earlier graphs. This graph confirms the others.

One thing I see that's different in this graph is the significant red spike at the right. Again, that is the Obama spending, but it didn't show up on the graphs in Part 3. (This stumped me until I remembered that the Part 3 graphs used MeasuringWorth GDP data, which stops in 2008. So my 5-year average ended in 2006.) Remove the last two data points here, and this graph is very similar to the other.

(I'm not ignoring the Obama spending for political reasons. I'm just making sure my data is valid.)


Some people say GDP is not a very good measure, because it doesn't count stuff they think it should count, and it does count stuff they think it shouldn't. Yeah, I don't know about that. I'm not an economist. I don't evaluate such views. But after doing the first three posts in this series, and now doing this one, I'm impressed with GDP. It accounts for price changes. It accounts for population changes. And it accounts for the changes that contribute to our standard of living. I'd say GDP is a pretty good measure.

<- Part 3

No comments: