[porphyrogenitus]

Peter Schiff was talking about recent changes in how GDP was going to be calculated by the USG last week.

I couldn’t tell if these were proposed changes, or if they’re already implementing them, and what impact this might have on GDP figures overall. I wonder if you have any comments on this.

[jmherbener]

The BEA has not yet changed the calculation of GDP.

http://www.bea.gov/newsreleases/national/gdp/gdpnewsrelease.htm

Here is there press release on the change.

http://www.bea.gov/scb/pdf/2013/03%20March/0313_nipa_comprehensive_revision_preview.pdf

The major changes are including R&D expenditures in GDP. According to the BEA this would have added $300 billion to the $14 trillion GDP in 2007.

Capitalizaton of entertainment and literary products. BEA estimates this would have added $70 billion to GDP in 2007.

Capitalization of ownership transfer costs of residential fixed assets. BEA estimates this would have added another $60 billion to 2007 GDP.

Switch from cash to accrual account of pensions. BEA estimates this would have added another $30 billion to 2007 GDP.

So, the $14 trillion GDP would have been $460 billion, 3.3 percent, larger in 2007. Because the BEA plans to revise the figures back to 1929, I don’t think the changes will drastically affect measured growth rates.

The BEA methodology has always been littered with difficulties such as those highlighted by the changes it plans to make.

http://www.bea.gov/methodologies/index.htm#national_meth

[porphyrogenitus]

Thanks tons for the fast and thorough reply! ^_^

Probably the impact will be to make the Debt-to-GDP ratio seem slightly smaller than it actually is, and, perhaps of more significance, make the government-spending-to-GDP ratio seem smaller than it was under the old measurement.

That and make international comparisons a matter of apples-to-oranges statistics, unless everyone else changes over to the “new, revised” method.

[samgheb]

On a slightly related topic I wonder about the significance of the changing in the calculation of the CPI. I asked Gary North about this and he said that what matters is the trend and not the actual number. John Mauldin has argued against John Williams Shadowstats as well on the same ground. Do you agree or does the actual number matter?

[jmherbener]

I’m assuming that by “number” you mean the CPI increased in the last quarter at an annualized rate of, say 2 percent. And by “trend” you mean, the CPI trended upward as it increased last year by 2 percent and this year it has increased by 5 percent at an annualized rate.

Because the CPI is a political football, the number matters. Imagine how different the political treatment of monetary policy would be if the official CPI numbers were as Shadowstats calculates them.

[Author]

Posts