Data issues and errors spread from global warming alarmism to economics…and from left to right

Whenever I mention the statistical chicanery, data manipulation, and errors behind “global warming,” my friends on the left go into high dudgeon. They can’t fathom that such things could occur among the “experts.” Well, this week, a new example of data issues (along with an error) popped up – in my discipline (economics), and from a paper largely used on the American and European right.

James Kirkup (Telegraph) explains the importance of the paper, known in academic-speak as Reinhardt and Rogoff, for the Conservative/Liberal Democratic Coalition in the United Kingdom:

“Rogoff and Reinhart” are Kenneth and Carmen, two economist (sic) whose influence over the Coalition’s economic policy is hard to over-state.

In essence, the economists argue that government above a certain level – 90 per cent of GDP – is catastrophically bad because it exerts a “significant negative effect on economic growth”.

Their argument, backed up with empirical data from lots of countries, played a major part in persuading Mr Osborne and his colleagues that the No 1 priority for the Coalition’s economic policy should be the reduction of the deficit and, ultimately, a check on UK government debt levels.

Mr. Osborne is George Osborne, the Britain’s Conservative Chancellor of the Exchequer, but Paul Ryan has also relied on the R&R paper as evidence to bring the American budget into balance, and it has also been part of the inspiration behind the Mediterranean being driven to Fauxsterity.

So when three economists at the University of Massachusetts (Thomas Herndon, Michael Ash, and Robert Pollin) tried to replicate R&R’s work – and couldn’t – they were naturally concerned. They were, however, able to access Reinhardt and Rogoff’s data…and what they found stunned them.

To wit….

  • Regarding data itself, Herndon et al found “data exclusions with three other countries: Australia (1946-1950), New Zealand (1946-1949), and Canada (1946-1950).” Now while one could argue that post World War II data should be excluded due to unusual circumstances, it does not explain why United States data was not excluded. As a result of the exclusion, Australia and Canada went from five years where the Debt/GDP ratio was above 90% to zero.  Meanwhile, New Zealand’s data went from five years (with an average of +2.8%) to one (with GDP growth at -7.9%). Speaking of the Kiwi data…
  • RR adopts a non-standard weighting methodology for measuring average real GDP growth within their four public debt/GDP categories. After assigning each country-year to one of four public debt/GDP groups, RR calculates the average real GDP growth for each country within the group, that is, a single average value for the country for all the years it appeared in the category.” Translation: Britain’s 19 years of data at high debt levels were given equal weight with New Zealand’s one year. I don’t think I need further explain the problem with this, especially given New Zealand’s prior data issues. It should be noted that Britain averaged 2.4% growth during the aforementioned 19 years. Belgium had 25 years of relevant data, but in addition to the weighting issue, the Belgian data was felled by…
  •  “A coding error in the RR working spreadsheet (which) entirely excludes five countries, Australia, Austria, Belgium, Canada, and Denmark, from the analysis.” As it happens, the first problem already dropped Australia and Canada, while Austria and Denmark had no relevant data anyway, but Belgium’s 25 years were excluded by the error.

So, when all of the above issues are backed out of the data and model, Herndon et al still find a reduction in economic growth when debt/GDP ratio hits 90%, but it’s a much smaller value than R&R found, and it turns out to be statistically insignificant. In fact, a regression analysis of the data found the inflection point (where things go from good to not-so-good) to be at 30% of GDP, not 90% (and that’s if one chooses to ignore the hideous R-squared statistic of 0.04 – on a scale of zero to one – which means the data explains hardly any of the changes in economic growth).

As it happens, my concern regarding government is more about its scope, size, and cost (in that order) than about how much it borrows. That said, many on the right have used R&R in part as justification for reducing deficits and debt. Based on the above, they may want to look to something else.

More to the point, this shows that climate policy isn’t the only political issue that has – to quote Coldplay – “castles stand…upon pillars of salt and pillars of sand.”

One Response to Data issues and errors spread from global warming alarmism to economics…and from left to right

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 56 other followers

%d bloggers like this: