Benn Steil


A graphical take on geoeconomic issues, with links to the news and expert commentary.

Print Print Cite Cite
Style: MLA APA Chicago Close


“Iceland’s Post-Crisis Miracle” Revisited

by the Center for Geoeconomic Studies
July 2, 2012

The Northern Lights are seen through a valley leading away from Iceland's Eyjafjallajokull volcano April 22, 2010. Lucas Jackson / Reuters


Back in July 2010, we produced a post examining the “Icelandic Post-Crisis Miracle,” as proclaimed by Paul Krugman.  We showed that Krugman’s “miracle” was merely an artifact of comparing changes in Iceland’s real GDP with that of Estonia, Ireland, and Latvia since the strategically chosen 4th quarter of 2007.

Why did Krugman choose the 4th quarter of 2007?  Because starting with any other quarter would have ruined his story.  Based on the GDP data available at the time he made his figure (which have since been revised), Iceland’s GDP had fallen a whopping 5 percentage points between Q3 and Q4.  By starting his story in Q4 Krugman managed to lop that off, making Iceland look much better.

We showed that the miracle story collapses as the starting date for the comparison is backed up.  What we find is a simple story of large booms and busts in Estonia and Latvia, and much smaller booms and busts in Iceland and Ireland.  Krugman’s post and our deconstruction are here and here.

Recently, Krugman has produced a slew of new posts reviving his claims in different forms.  Geo-Graphics readers have, not surprisingly, asked us to revisit the question of Iceland’s economic performance.

Krugman produced this figure in a post on June 14, showing that as of the 1st quarter of 2012 Iceland had a better real GDP growth performance relative to its GDP peak than the three Baltic states (Latvia, Estonia, and Lithuania) and Ireland had relative to their various GDP peaks.  “Looking at this,” Krugman asks rhetorically, “would you have expected that Latvia would be lionized as the hero of the crisis?”

The answer is, of course, “no.” But Latvia only looks so bad because Krugman chooses to tell his story about post-crisis performance only in terms of how each country has performed since its peak. This makes little sense in the context of the IMF’s 2009 staff report which concludes that Latvian “output exceeded potential by 9 percent” in 2007. (Updated 7/3/2012 2:04 p.m.)

What if we decided to tell the story only in terms of how they have performed since their troughs – that is, how well they have recovered since they hit bottom? Here it is:

We don’t yet have the Eurostat GDP data for Iceland in the first quarter of 2012, so we’ve plugged in data from the Icelandic government (as Krugman must have done). As can be seen, Iceland’s performance has only been on par with the bottom of the Baltic pack, Lithuania. Latvia’s performance has been better, and Estonia’s markedly better.

Once again, Krugman has relied on a Potemkin-Village graphic to illustrate his wider claim, which is that Icelanders derive unambiguous net benefits from their government obliging them to hold and transact in a national currency that their trading partners will not accept. (80% of Greeks consistently reject going back to such a state.)

Below we update the figure in our July 2010 Geo-Graphic, comparing Iceland’s economic performance with that of Estonia, Ireland, Latvia, and Lithuania going back to 2000.

Iceland comes in tied for last with Ireland (for which 2012 Q1 data are not yet available) – well behind Lithuania, Estonia, and Latvia. Since 2009, those three have been growing strongly, while Iceland and Ireland have largely stagnated.

Sorry, Virginia, there is no Icelandic miracle.

Geo-Graphic: Post-Crisis Iceland: Miracle or Illusion?
Krugman: Peripheral Performance
Financial Times: Iceland: Recovery and Reconciliation
Reuters: Baltic Countries’ Austerity Lesson for Europe—Just Do It

Post a Comment 23 Comments

  • Posted by Mark Smith

    Why would you chose to measure from the bottom of the trough? If policies were put in place that encourage growth, prior to a country hitting the bottom of the trough, it would be reasonable to assume that GDP wouldn’t have dropped as much as in a country where pro-growth policies were not put in place, so the pro-growth countries would have a shallower trough. It would also seem that in general, growth rates will be higher when climbing our of a deep trough relative to a shallow trough. Don’t we really want to know where a country is relative to it’s potential GDP? And wouldn’t measuring from the peak be a better way to measure that?

    From the last graph, it looks like Iceland is closer to getting back to it’s trend GDP than is Latvia, Lithuania, or Estonia, isn’t that more important than where they are relative to the bottom of the trough?

    And yes, the Baltic states have had faster growth than Iceland, but isn’t that most likely due to starting at a much lower standard of living? China has been maintaining growth rates much higher than the the advanced economies, but I doubt many people would suggest adopting most of their economic policies.

  • Posted by Matt Waters

    This post completely ignores the real factor in Keynesian/Monetarist analysis: unemployment.

    Keeping AD high through monetary or fiscal stimulus does not prevent something bad from happening. If a country like Iceland has a grossly inflated financial sector and property bubble, that financial bubble will still hurt Real GDP. What should not follow is that a “general glut” where the property bubble causes an additional hit to Real GDP through higher unemployment and idleness.

    On this score, Iceland has 6.6% unemployment while Ireland has 14.2% unemployment and Latvia has a 15.2% unemployment. The large divergence in unemployment rates happened for some reason and it is not like Ireland had a much larger property/financial bubble than Iceland. Instead, the only reasonable explanation is that Iceland kept AD high through depreciating their currency.

  • Posted by joddeHaa

    Your being disingenuous by implying that Krugman is trying to hide something by comparing peaks rather than troughs. It is in fact standard practice. Here he is commenting on that exact complain earlier:

    “””here’s the point: the biggest single influence on revenue is the state of the business cycle (which is why austerity in the face of a liquidity trap is so ineffective even at reducing deficits). And in normal times the business cycle depends much more on what the Fed does than on anything the rest of the government does (again, it’s different when there’s a liquidity trap).

    So what you want is to somehow abstract from the business cycle. And the easiest way to do that is to compare business cycle peaks, times when the economy is at more or less full employment. True, employment is fuller at some peaks than at others, but those differences are small, whereas the depth of the slump at recession troughs is much more variable. Hence, Anna Karenina.

    The peak-to-peak interpolation thing is standard practice in many analyses; it’s how the Fed estimates capacity utilization, it’s how the Social Security Administration makes long-term projections, and so I’ve used”””

  • Posted by Basho

    With this kind of analysis, it is better to have been laid off from a $30.00 an hour job, hired at $9.00 an hour, and then given a raise to $15.00 an hour than to be someone that had their pay frozen at $30.00 an hour The guy earning $15.00 an hour has had substantial improvement since trough!

    As for the last graph, the basic lesson is that countries with lower levels of GDP per capita tend to experience faster growth. Iceland and Ireland have a GDP per capita (PPP basis) nearly twice that of Estonia and more than twice that Lithuania and Latvia.

  • Posted by orionorbit

    There is a reason Krugman uses top-to-top comparisons and that is because this is an accepted measure of comparing potential Outputs. On the other hand, your choice to measure performance the other way around is completely ad hoc as the different degrees of price stickiness among countries blur the picture.

    Your argument is VERY flawed in another respect as well. If I make 100 bucks, then i go down to making 75 and then 80 and I’m projected to make 2 next year, I am doing much worse than someone who was making 100 bucks, got down to 85, then 88 and is projected to make 89 next year.

    Additionally, last time I checked governments cared about unemployment and in this sense Latvian recovery has been abysmal as its employment is way way bellow potential. Not the case with Iceland.

    I agree that there is no miraculous recovery in Iceland but there is some kind of recovery, while the Latvians have to tolerate much higher unemployment.

  • Posted by Sanman

    Thanks for posting this. A graph should illuminate the answer to the question. I thought the question in this case is “which policies, Iceland’s or Latvia’s, were most helpful in keeping up economic performance after the bubbles popped?” That is, which policies best protected people from suffering?

    In that case, tracking performance relative to the peak makes the most sense. Using the trough would answer “what helped economies bounce back best after the disaster reached its maximum?”, which is not helpful since the point is to avoid falling into a deep trough.

    In the second set of graphs, the overall performance of the economies since 2000 is beside the point about the bubble, since they’re different countries whose economies differ in many ways. Of course they have different overall rates of growth.


  • Posted by Pedro Dias

    Two points: Iceland was a relatively wealthy country in 2000, which makes your final point at best a fuzzy one; and Iceland’s “losses” included a large, largely imaginary, banking and financial sector, which will never, and arguably should never, recover.

  • Posted by R plaster

    Dr. Krugman is fully capable of defending himself and has done so. What I’m curious about is why do these articles not have an identifiable author? Individual responsibility rather than collective anonymity is to my mind what responsible conservatism is all about.

  • Posted by Jamison

    The last graph confounds the long-run growth of these countries and the bubble/collapse’s effect on their economies. I do not know anything about economics but I can see that. Iceland is the most developed country of the group so its potential GDP growth would be lower than the rest, right? Ireland is next and then the Baltics would have the highest normal GDP growth under normal conditions. Don’t you need to control for that?

  • Posted by Shameer

    What you fail to note is that Ireland and Iceland are first world countries with incomes similar to that of the US, while the Baltics are middle income countries with per capita nominal incomes roughly a third of the other two (

    Since middle income countries grow faster on average, comparing the growth since 2000 is meaningless. Also, comparing since the trough is less than useful since you are penalizing good economic management that averted a deeper trough.

    Although it’s not perfect, comparing from the peak is the best of the three methods.

  • Posted by Gustav Schmoller

    What about unemployment? Unemployment rose much more dramatically in the Baltics than in Iceland — and it has fallen much faster in Iceland than in Latvia — even though the Baltics have been major “exporters” of unemployed labor.

  • Posted by Seth S

    Perhaps there is no Icelandic miracle, but then again is it fair to compare against a trough? If the Baltics theoretically stopped their economies for a year or suffered a huge depression that would say, show a 50 or 75% reduction in year over year GDP (or in your revised chart, go back to Q1 2000 GDP) and then grow to show an increase of 50% from the Q1 2000 numbers, that would look great…. but if they had been as high as 60 or 70 growth from the 2000 Q1 numbers, there is still a hole (even accounting for a “bubble” attributing to the absolute peak)

  • Posted by Marques Mendes

    Guys you should not compare long term performance in relation to a specific year but to an average. Using peaks or troughs is only useful when assessing falls and recoveries. As simple arithmetic will show if your fall is 50% you need a rise of 100% to get to your previous position. As the popular say goes the deepest the hole the more you need to climb.

  • Posted by nemonullnichts

    This is one of the less successful arguments I’ve read on the site. It appears to overstate another party’s argument by quibbling (unjustifiably?) with their methodology and while recapitulating it.

    First, shouldn’t we expect the less thoroughly developed economy to see gains of greater magnitude precisely as is illustrated in the latter two graphs? In other words, could we agree that all parties err in comparison of Latvia, Lithuania, and Estonia and should hew to the closer comparison of Ireland, or perhaps Spain?

    Second, it seems that the most useful way to focus on post-crisis progress would be a peak to trough to peak comparison. Otherwise we could be left with mistaken impression that a 12% recovery following a 31% loss is a greater achievement than an 11% recovery following a 17% loss.

    It’s true that Latvian GDP was bolstered by an artificial stimulus in the run up to the crisis. However, so too were most economies in this time period, whether by domestic fiscal policies or unsustainable capital influx from abroad. Neither Iceland nor Ireland could have achieved their GDP peaks through manic financial or real estate sectors through their domestic demand and investment resources alone.

    I think many of CFR’s readers would appreciate if they elaborated on the Icelandic policy’s serious consequences to trade and other international cooperative efforts instead. While it’s disputable that, “…their government oblig[es] them to hold and transact in a national currency that their trading partners will not accept,” given that the determination was made democratically, few parties would be as well-positioned or as well-equipped as the commentary pages on this site.

  • Posted by Antonio Casal

    I thought PK was exaggerating when he termed this post “remarkably stupid,” but unfortunately he is right. Looking at where each country stands relative to its prior peak is apples-to-apples (you get no credit for having bounced a bit more from the worst bust, as Latvia clearly has). BTW, it’s also what matters most to the average person–how close am I to attaining the living standard I enjoyed when this whole mess started?
    The long-run comparison of poor post-communist countries with Ireland and Iceland makes no sense. And lastly, Krugman never said “Icelanders derive unambiguous net benefits from their government obliging them to hold and transact in a national currency that their trading partners will not accept.” What he said was that devaluation and burden-sharing by bank creditors was a better policy for facing a financial bust than austerity, retention of the currency peg and a refusal to make bank creditors take losses.
    It amuses me that you cite the Greek public’s support for staying in the euro as evidence of the wisdom of that policy. On what other topics do you give so much credit to the Greek street?

  • Posted by Miguel Sanchez

    I think the key to this debate is what constitutes ‘success’ in the first place. If you’re a Keynesian (or a Stalinist) and maximising the volume of national output counts as ‘success’ then Iceland seems to shape up.

    But we need to look at the impact on balance sheets as well, as I suspect (but am too lazy to investigate) that the experiences are vastly different. Since Iceland went down the currency depreciation route – and since they had a preponderence of foreign-currency mortgages – they can probably look forward to devoting much more of their future income to paying off the debt they racked up during the go-go years.

  • Posted by Jacob

    The weakness of your analysis here – and Krugman’s as well; I’m not nearly masochistic enough to try defending his argumentation – is that you rely solely on GDP statistics. This is seems problematic, if for no other reason than that competent, self-coherent advocates (*cough*) of Iceland-style paradigms tend, in my experience, to readily eschew GDP growth as the primary, certainly sole, measure of successful economic function.

    I won’t belabor the mostly well-known and obvious justifications for this perspective – but I’m interested in whether, adopting a different or broader choice of metrics for the sake of investigation, we would find a more interesting and less straightforwardly unexceptional picture of Icelandic policy impacts.

    Any interest in appending something like this?

  • Posted by Nardus

    This does not expose Krugman using an artifact just to make his point, but rather a shameful lack of understanding of the way such phenomena are analysed in economics on the part of the CFR.

  • Posted by Peter Kropf
  • Posted by Patrick Signoret

    Everybody take a deep breath and look at all three graphs. What do they show? THEY ALL SHOW THE SAME THING.

    With the crisis, Latvia’s economy fell much more than Iceland’s did. Since then, Latvia’s recovery has been relatively much weaker than Iceland’s.

    Again: taking different base years doesn’t matter (as long as you show enough history). All three graphs show the same thing.

    Which graph shows it most clearly? The one with 4Q07=100. So that’s the best graph to use to illustrate the point. That’s why Krugman used it. His view is supported in all graphs, but it’s more clear in the one he chose.

    The graph with trough=100 is ridiculous because countries that were hit harder will of course bounce back more quickly. But that’s ok, because the graph shows history, so one can still see that, just to be even with Iceland, Latvia’s recovery should have been much stronger.

    The graph with 2000=100 is okay; it shows that Latvia is a faster-growing economy (because it’s poorer). Doesn’t change the fact that it’s recovery has been proportionally much worse than Iceland’s.

  • Posted by Charles Munger

    Doesn’t a twenty-three percent drop in eighteen months bother anyone?

  • Posted by not scott sumner

    Scott Sumner: I’ve never followed Krugman’s practice of looking at how we are doing in comparison to the peak of the previous business cycle, which was late 2007. In most cases a cyclical peak is above the economy’s natural rate of output, i.e. the economy is somewhat overheated.

  • Posted by Rondreis IJsland

    Nice to read about Iceland’s issue’s. It is a beautiful island overall.

Post a Comment

CFR seeks to foster civil and informed discussion of foreign policy issues. Opinions expressed on CFR blogs are solely those of the author or commenter, not of CFR, which takes no institutional positions. All comments must abide by CFR's guidelines and will be moderated prior to posting.

* Required