Zombie ideas are much scarier than actual zombies. For one, zombie ideas are real. For another, nothing can kill them. Not even a bullet to the intellectual head.
Just look at the austerians.
It was only a week ago that the cult of deficit-cutting was in quick retreat. Harvard professors Carmen Reinhart and Ken Rogoff, the authors of the austerity ur-text
to making a number of errors, including one with a spreadsheet formula, that made their seemingly dramatic results much less so. As Owen Zidar
points out, growth doesn't hit some kind of tipping point when public debt hits 90 percent of GDP, as Reinhart and Rogoff claimed, but rather falls at a steady rate between debt levels of 50 and 200 percent of GDP. A bit of simple math, which Brad DeLong
goes through, shows that this relationship between higher debt and lower growth isn't anything approximating an argument against more stimulus. As long as interest rates are zero and unemployment is high, there's a strong case for doing more.
In other words, L'Affaire Excel
has once again left austerity as a policy without justification. Now, the spreadsheet error wasn't that significant, but it called attention to the bigger one
: Reinhart and Rogoff oversold
correlation as causation. That's become harder to do now that the correlation has become weaker. Even Euro Commissioner Olli Rehn, who thought austerity would work if those pesky Keynesians
would just stop pointing out that it wasn't
, has given up on it. Here's how Rehn framed the budget debate back in March
And it is widely acknowledged, based on serious research, that when public debt levels rise about 90 percent they tend to have a negative economic dynamism, which translates into low growth for many years.
And here's what Rehn said about it last week
The precise causal relationship between debt levels and growth is a complex one. There is no hard-and-fast rule; it is affected also by many country-specific factors.
But this ray of realism aside, the real
high priests of austerity have not given up on fiscal rectitude as the one true path to economic salvation. Those are the austerians at the European Central Bank (ECB). As Ryan Avent
of The Economist
points out, euro-debtor countries can only spend as much as the ECB lets them. The ECB pushed down their borrowing costs last year by explicitly backstopping them, but it can just as easily push them back up by removing the backstop -- which it will do if they backslide on deficit-cutting. Indeed, ECB Executive Board member Jörg Asmussen
was back to touting ... Latvia
as a shining example just last week:
Latvia is the leading example of how to adjust through internal devaluation, and it is a model for others in the euro area. Its "V-shaped" economic recovery illustrates what can be done with a strong consensus to undo the excesses of the past. After an initial fall in GDP of almost 18% in 2009, GDP increased by more than 11% from 2010 to 2012, and unemployment has fallen by almost 7 percentage points from its peak.
Latvia followed a familiar script: boom, bust, and bailout. At its peak in both 2006 and 2007, Latvia was borrowing over 22 percent
of its GDP from abroad -- a private-sector binge that fueled a massive housing bubble. When the money music suddenly stopped after Lehmangeddon, this fragile web of debt imploded, and with it, the entire economy. Latvia only managed to avoid national bankruptcy by going to the EU and IMF for a bailout
, which allowed it to keep its currency pegged to the euro -- but at the cost of promising deep austerity. This combination of tight money and tight budgets sent Latvia into a historic slump: GDP fell over 25 percent, and unemployment hit 20.7 percent.
But Latvia's depression isn't quite so great anymore. Its economy has grown 5.5 percent each of the past two years, and unemployment is just under 14 percent. Is this proof that there can be gain after pain? Yes. Is it vindication? No. Economically, it never made sense for Latvia to keep its currency peg (though it did politically due to Russophobia) or to cut spending amidst a slump. As Paul Krugman
point out, Latvia's GDP is still far below its pre-crisis peak, and even more so than exemplars like Greece, Ireland, Italy, Portugal, and Spain are below theirs. It's not that much better on the jobs front. Now, Latvia does
have much lower unemployment than Greece and Spain
, but that's in part due to how many people have left Latvia
. As you can see below, Latvia only recently passed Greece and Spain in total jobs relative to their peaks.*
(Note: Latvia rebenchmarked its data in 2011, so I had to adjust the numbers for 2012. I took the numbers from Latvia's Central Statistical Bureau, calculated the quarterly percent changes in 2012, and used those to project what the 2012 numbers would have been under the old system. All other data is from Eurostat and St. Louis FRED).
There was never much to learn from Latvia. As Martin Wolf
points out, it's a small, open economy of two million people with lots of foreign-owned banks it never has to bail out. In other words, it's not exactly replicable. But why would we want to replicate it? Deep cuts to public sector wages and spending in Latvia turned a severe recession into a deep depression. Its "successes" today only highlight how far it fell. GDP is growing quickly, in part, because of how low its base was after the slump; and unemployment is falling quickly, in part, because of how many people emigrated during the slump. Meanwhile, the U.S., which did do countercyclical fiscal and monetary stimulus, didn't fall nearly as far, and has just as much, relative to the trough.
But zombie ideas do not die, especially in the euro zone. Nothing will convince the ECB that Latvia isn't a success story -- unless it's more convenient to claim Estonia is instead. Regardless of the fact, the ECB will keep pushing southern Europe into perma-slump as the path to prosperity.
Now that's scary.
* The original version of this story mistakenly used the re-benchmarched data for Latvia from Eurostat, instead of the original 2011 numbers and adjusted 2012 numbers. I got the new, unbenchmarked numbers from Latvia's Central Statistical Bureau, but they only go back to 2002, instead of 2000, so I cut back my sample by two years. I also removed Lithuania from the sample, because I was worried about the possibility of a similar benchmarking issue.