BUSTED: Paul Krugman removed 20 yrs of data from chart to show correlation that wasn't really there

Business Insider ^ | 05/21/2015 | CULLEN ROCHE, PRAGMATIC CAPITALISM

[Mr. K]

There are 3 types of lies: Lies, Damnable Lies, and Statistics.
-Mark Twain

[Mr. K]

Lies, Damned Lies, Statistics, Democrat Talking Points.
-Mike Kosak

[fhayek]

I approach economics with logic and common sense. I reason, how can you borrow and print money forever and expect a healthy and well-functioning economy. People like Krugman argue things like this with their charts and their statistics, but I just can't get past the basic illogic of it

[Redmen4ever]

Krugman has done his best to degrade the Nobel Prize in Economics. I will, however, point out that his contributions cited in his award concerned international trade (and are fine). His macroeconomics are whack and often involve thoroughly discredited ideas and techniques. Within economics, it is well known that serial correlation in the data leads to spurious correlation over “short intervals.” Indeed, the “length” of an interval is not defined by the frequency with which the data are revealed, but by the time it takes to observe a new datum relatively unaffected by the time series own past data. For many time series, it is appropriate to work with first differences. Several economists have been awarded for the contributions to the statistical analysis of time series, including Clive Granger and Christopher Sims.

[tanknetter]

I thought I read somewhere that Ktugman was run out of Princeton by Paul Volker

[dead]

Exactly, but it is all irrelevant to the point of this article - which is that Krugman manipulated the available data set to fit his narrative. That behavior is pretty unethical.