Boom, bust, recovery: forensics of the Latvia crisis.

Blanchard, Olivier J. ; Griffiths, Mark ; Gruss, Bertrand 等

VI.B. Unemployment

As of June 2013, the unemployment rate was still high at 11.4 percent. The question is how far this rate was from the natural unemployment rate, and thus how much remained to be done. To think through this question, we have two--admittedly imperfect--tools: the Beveridge curve and the Phillips curve.

Figure 16 plots the Beveridge curve, which shows the relation between the unemployment rate and the vacancy rate, from 2005Q1, the first quarter for which data on vacancies (from an official survey) are available. After an early inward shift of the curve toward the end of the boom, the relation appears quite stable. Indeed, as unemployment starts declining from its peak, it appears to be retracing its movement in the slump. In short, there is no evidence of an adverse shift in the Beveridge curve due to the crisis. (38)

However, this does not settle the issue of what the natural rate might be. For this, we must look at the relation between inflation and unemployment, the Phillips curve. Figure 17 plots core inflation minus expected inflation against the unemployment rate, as well as the corresponding regression line. (Details of the estimation are given in the appendix to this paper and in the online appendix as well.)

The point estimate for the unemployment rate at which core inflation is equal to expected inflation is a surprisingly high 13.3 percent (a 95 percent confidence interval ranges from 11.7 percent to 14.8 percent). The figure also makes clear that an unemployment rate below 8 percent has been typically associated with large increases in inflation. This is indeed what we found earlier when looking at inflation in the boom. Thus, using this metric, the actual unemployment rate is rapidly approaching the natural rate.

[FIGURE 16 OMITTED]

This raises a final question, which, while not central to the issues of this paper, is nevertheless intriguing: How can a country with a low minimum wage, weak unions, and limited unemployment insurance and employment protection have such a high natural rate? We do not have a good answer. (39) High unemployment appears to reflect high duration rather than high reallocation and high flows through the labor market. The "Lilien index," defined as the standard deviation of sectoral employment growth rates for 10 sectors for the decade 2001-10, is substantially higher in Latvia than in Germany or France. But job turnover, defined as the sum of job creation and job destruction, appears to be similar in Latvia to that in Germany and France (Boeri and Garibaldi 2006). Unemployment duration appears similar to that for the European Union, with a large upper tail. This might be caused by high reservation wages due to a still extensive informal economy (including home production and barter networks, especially in rural areas) or to skill mismatches.

[FIGURE 17 OMITTED]

To summarize: The actual unemployment rate is probably close to the natural rate of unemployment. Latvia may well want to take measures

to reduce its natural rate, but the recovery from the slump is largely complete.

VII. Conclusion and Tentative Lessons

The Latvian boom-bust recovery story is a striking one. While we have tried throughout this paper to stick with the facts and avoid normative statements, here we will venture to draw some lessons, beginning with six that are more narrowly focused and then addressing larger matters.

The first lesson is an old one: Healthy booms often turn unhealthy, and policymakers often react to them too late. The boom in Latvia was healthy until 2005, unhealthy thereafter. Clearly Latvia could not have avoided the adverse effects of the global crisis, but had monetary policy been tighter from 2006 on, the adjustment would have been easier.

The second lesson is that if a country's financial sector is largely composed of foreign subsidiaries, it is a good idea for its government to be friendly toward the parent banks. Had Nordic banks not largely maintained funding for their Latvian subsidiaries, the outcome for Latvia could have been much worse.

The third lesson is that strict currency boards do not mix well with sudden stops. If the Latvian Treasury and central bank had followed strict currency board rules and not provided funding to the banks, the decline in output would likely have been much larger.

The fourth lesson is that productivity gains can sometimes happen quickly. The general assumption in European adjustment programs was that competitiveness would first be achieved by wage cuts and that productivity increases, spurred by structural reforms, would come over time. This assumption has turned out to be largely correct in the southern periphery countries, where productivity gains have so far been limited. However, in Latvia productivity increases were sufficient to require only a limited adjustment of nominal wages.

The fifth lesson is that, even if a decrease in unit labor costs leads to an increase in profit margins rather than a decrease in prices, the increase in profit margins can sometimes lead to a rapid increase in exports.

The sixth lesson is that, for political purposes, the growth in output may matter as much as or more than the output level. One of the striking aspects of the Latvia story is that, while most European governments lost to their opposition during this period, Latvia had the same prime minister from March 2009 to December 2013. It is true that he could blame the previous governments for the fall in output, but for most of his own tenure unemployment was very high, and still he remained in power. The fact that growth was positive, even if output was low, was seen as a sign of success. (40)

We turn to larger issues when we consider the questions that the debate about Latvia has focused on with greatest intensity, namely the two aspects of its adjustment program: front-loaded fiscal consolidation and the choice of internal vs. external devaluation.

The first issue is the large front-loaded fiscal consolidation that the government enacted. As our analysis showed, the timing of events makes it clear that fiscal adjustment was not responsible for much of the drop in output. Moreover, much of the fiscal adjustment coincided with a return to growth in Latvia. But these facts do not settle the question of whether fiscal consolidation had adverse effects on demand and output; as we saw, much of the output collapse was due to a credit crunch and to the option value of waiting. As both dissipated, the impulse for growth may well have been strong enough to offset the adverse direct effects of the consolidation.

There is some evidence that the announcement of a clear fiscal path was associated with increased confidence, reflected in lower CDS spreads. But the evidence also suggests that the large decrease in the rates relevant to private borrowers was associated with the strengthened credibility of the peg and the lowering in exchange rate risk. Whether the front-loading aspect of the fiscal adjustment made the whole adjustment program more credible cannot be settled. Adjustment in the 2011 budget was minimal: this may be interpreted as adjustment fatigue, justifying the earlier front loading; or it may be interpreted as coming from the perception that, given front loading, enough had been done already. In short, the experience of Latvia sheds little light on the issue of the optimal speed of fiscal consolidation.

The second issue concerns the choice to pursue an internal rather than an external devaluation. It is fair to say that this approach worked better than most analysts had expected. The main argument for an external devaluation--an adjustment of the nominal exchange rate--is that it would have worked faster; but the internal devaluation worked surprisingly quickly as well. It did not work in quite the way one might have predicted, however. As we showed, much of the improvement in the tradables sector came from productivity increases rather than nominal wage decreases. And much of the decrease in unit labor costs was reflected in an increase in profit margins rather than a decrease in prices. Still, exports increased rapidly, pulling the recovery. One might even argue that a nominal devaluation might have led to less pressure on firms to increase productivity.

Do these lessons extend beyond Latvia? The evidence from adjustments in the southern euro periphery countries suggests great caution. Frontloaded (although less so than in Latvia) consolidation has been associated there with negative growth, and internal devaluations have been associated, at least initially, with labor hoarding and decreases rather than increases in productivity. The challenge is to identify what factors in periphery countries make them different from Latvia, and whether there are other countries that share characteristics with Latvia (such as a small open economy or a record of rapid productivity growth) that could follow a similar approach in the future.

And finally: Was it a wise decision for Latvia to join the eurozone in 2014? Based on the evidence, the answer is that the case for Latvia joining is strong, perhaps even more so than for some of the existing members. Small open economies are very sensitive to capital flows. Floating can lead to large movements in the exchange rate. One option is to reduce capital mobility. The other is to peg. Pegging may be consistent with sufficient real exchange flexibility if prices, wages, productivity, or migration can adjust fast enough, as appears to have been the case for Latvia. And if Latvia is going to peg, it is probably better off becoming a member of the euro and thus having access to a lender of last resort. However, once the catching-up process has run its course, if Latvia is faced with shocks of similar magnitude in the future, it may suffer from the same problems of adjustment that some of the current euro members face today.

APPENDIX

CONSTRUCTION OF THE EMPLOYMENT SERIES The usual choice for employment data is the Labor Force Survey (LFS). However, Latvia's LFS data are being revised (together with population data) and show a sharp discontinuity in 2011, when the national population and housing census indicated that Latvia's population was around 10 percent lower than previously assumed. This invalidated the earlier extrapolation of LFS sample data to national aggregates for the period between population censuses. Latvia's Central Statistical Bureau has still to revise 2000-2011 LFS data to correct for this population gap.

Normally, problems like this can be solved by multiplicative splicing, in this case using the revised data for 2011. This would suggest a 12.4 percent drop in employment from 2008Q2 to 2013Q2. However, that approach would still be problematic, since working backward would imply a population 10 percent lower than recorded in the 2000 census. The decline in population--and the resulting adjustment of LFS statistics--needs to be apportioned correctly over time, and the decline might not be smooth but instead concentrated during the crisis, when emigration was greatest.

For this reason, in this paper we use data on the number of employees or "occupied posts" as our measurement of employment to break down the ULC correction into wages and productivity. These data are obtained by the national statistical office by surveying enterprises and government institutions and are not affected by the recent correction in population estimates. (41) While the data have some drawbacks (for example, they do not include those that are self-employed), we believe they are a more reliable measure of employment until the LFS data are revised.

PHILLIPS CURVE ESTIMATION AND THE NATURAL RATE The relation between inflation and unemployment shown in figure 17 is based on estimates from a simple Phillips curve-type regression. The model specification is given by:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

where [[pi].sub.t], is the year-on-year inflation rate of core consumer prices (that is, excluding energy and unprocessed food components); [[pi].sup.e.sub.t] is expected inflation; [u.sub.t] is the seasonally adjusted unemployment rate (from Eurostat based on LFS data); and [u.sup.*] is the (constant) rate of structural unemployment consistent with no inflation pressure (NAIRU). Additional variables to control for supply shocks that can affect inflation dynamics beyond the effect of domestic demand, such as changes in indirect taxes and inflation in the relative prices of imports, were considered, but they were discarded as they turned out to be insignificant in this specification. (42)

Expected inflation ([[pi].sup.e.sub.t]) is constructed as the average of: (i) past inflation in Latvia (the average over the last 4 quarters); and (ii) core inflation in the euro area (year-on-year, average over the last 4 quarters) plus 1 percent-- assuming some catch-up of price levels in Latvia as income levels converge toward the euro area average. The weights of these components are estimated but are constrained to sum to one.

The estimation results are presented in table A.1. The estimate for the weight of lagged domestic prices in expected inflation ([delta]), used to construct the difference in inflation plotted in figure 17, is 0.57 and is significant at the 1 percent level (its Newey-West corrected standard error is 0.11). The point estimate for the NAIRU ([u.sup.*]) is 13.3 percent. A 95 percent "Gaussian" confidence interval (constructed by the approach suggested in Staiger, Stock, and Watson 1996) for [u.sup.*] ranges from 12.2 percent to 14.3 percent (using the Delta method, instead, the 95 percent confidence interval goes from 11.7 percent to 14.8 percent).

Table A1. Phillips Curve Regression

 Dependent variable:
 [[pi].sub.t]

 Standard
Variable Coefficient error (a)

Constant 7.30 1.51
Weight in [[pi].sup.e.sub.t] ([delta]) 0.57 0.11
Unemployment ([beta]) -0.55 0.11

NAIRU (-Constant/[beta]) 13.26
95% confidence interval
 Gaussian (b) 12.2-14.3
 Delta method 11.7-14.8

[R.sup.2] 0.79
Obs. 62

 Dependent variable:
 [[pi].sub.t]

Variable t-statistic Probability

Constant 4.83 0.00
Weight in [[pi].sup.e.sub.t] ([delta]) 5.04 0.00
Unemployment ([beta]) -4.98 0.00

NAIRU (-Constant/[beta])
95% confidence interval
 Gaussian (b)
 Delta method

[R.sup.2]
Obs.

(a.) Newey-West corrected standard errors.

(b.) Based on approach suggested in Staiger, Stock, and Watson (1996).

ACKNOWLEDGMENTS We thank Shekhar Aiyar, Anders Aslund, Bas Bakker, Rudolfs Bems, Aaron Eglitis, Mihails Hazans, Morten Hansen, Martins Kazaks, Joong Shik Kang, Alvar Kangur, Alexander Klemm, Ayhan Kose, Olegs Krasnopjorovs, Nicolas Magud, David Moore, Christoph Rosenberg, Uldis Rutkaste, Sergejs Saksonovs, Gabriel Srour, the editors, and our discussants for suggestions and comments (and also for teaching us about Latvia), and Bartek Augustyniak and Andreas Schaab for research assistance. All three of us (the authors) are at the International Monetary Fund (IMF). Olivier Blanchard was the IMF's chief economist while the IMF- and EU-supported program was designed and implemented. Mark Griffiths was the IMF mission chief to Latvia from 2009 to 2012. Bertrand Grass was a member of the IMF's Latvia mission team from 2012 to 2013.

References

Anosova, Daria, Konstantin Sonin, Alf Vanags, and Anna Zasova. 2013. "Structural or Cyclical? Unemployment in Latvia since the 2008-09 Financial Crisis." CEPR Discussion Paper no. 9525. Washington, D.C.: Centre for Economic Policy Research.

Ashcraft, Adam. 2005. "Are Banks Really Special? New Evidence from the FDIC-Induced Failure of Healthy Banks "American Economic Review 95: 1712-30.

--. 2006. "New Evidence on the Lending Channel." Journal of Money, Credit and Banking 38, no. 3: 751-75.

Aslund, Anders, and Valdis Dombrovskis. 2011. "How Latvia Came Through the Financial Crisis." Washington, D.C.: Peterson Institute for International Economics (May).

Bakker, Bas, and Christoph Klingen, editors. 2012. "How Emerging Europe Came Through the 2008/2009 Crisis." Washington, D.C.: International Monetary Fund.

Bayoumi, Tam, and Ola Melander. 2008. "Credit Matters: Empirical Evidence on U.S. Macro-Financial Linkages." Working Paper no. 08/169. Washington, D.C.: International Monetary Fund.

Bems, Rudolfs, and Julian Di Giovanni. 2013. "Sudden Stops and Expenditure Switching: Scanner Data Evidence from the 2008-09 Global Financial Crisis." Mimeo. Washington, D.C.: International Monetary Fund.

Betjkovskis, Konstantins. 2012. "Competitiveness of Latvia's Exporters." Working Paper no. 3/2012. Riga, Latvia: Bank of Latvia.

Bernanke, Ben, and Cara Lown. 1991. "The Credit Crunch." Brookings Papers on Economic Activity 2: 205-47.

Berrospide, Jose, and Rochelle Edge. 2010. "The Effects of Bank Capital on Lending: What Do We Know, and What Does It Mean?" International Journal of Central Banking 6, no. 4: 5-54.

Blanchard, Olivier. 2007. "Adjustment within the Euro: The Difficult Case of Portugal." Portuguese Economic Journal 6, no. 1 (April): 1-22.

--, Mitali Das, and Hamid Faruqee. 2010. "The Initial Impact of the Crisis on Emerging Market Countries." Brookings Papers on Economic Activity, Fall.

Blanchard, Olivier, and Lawrence Katz. "Regional Evolutions." 1992. Brookings Papers on Economic Activity, Spring: 1-75.

Boeri, Tito, and Pietro Garibaldi. 2006. "Are Labour Markets in the New Member States Sufficiently Flexible for EMU?" Journal of Banking and Finance 30, no. 5: 1393-1407.

Biggs, Michael, Thomas Mayer, and Andreas Pick. 2009. "Credit and Economic Recovery." DNB Working Paper no. 218. Amsterdam: Netherlands Central Bank.

Claessens, Stijn, Ayhan Rose, and Marco Terrones. 2009. "What Happens during Recessions, Crunches and Busts?" Economic Policy 24: 653-700.

Calomiris, Charles, and Joseph Mason. 2003. "Consequences of Bank Distress during the Great Depression." American Economic Review 93, no. 3: 937-47.

Central Bank of Ireland. 2011. "Compositional Effects in Recent Trends in Irish Unit Labour Costs." (Box A on pages 22-24, prepared by Derry O'Brien.) Dublin.

Dao, Mai, David Furceri, and Prakash Loungani. 2013. "Regional Labor Adjustment in Currency Unions: Evidence from Europe and the United States." Presentation at the 53rd European Regional Science Association (ERSA) Congress, Palermo, Italy, August 27-31. http://www.ersa.org/53rd-ersa-congress/.

Di Comite, Francesco, Gabriele Giudice, Radoslav Krastev, and Daniel Monteiro. 2012. "Fiscal Consolidation in the Midst of the Crisis." EU Occasional Paper no. 120. http://ec.europa.eu/economy_finance/publications/occasional_paper/ 2012/op 120_en.htm.

Driscoll, John. 2004. "Does Bank Lending Affect Output? Evidence from the U.S. States." Journal of Monetary Economics 51, no. 3: 451-71.

Duchin, Ran, Oguzhan Ozbas, and Berk Sensoy. 2010. "Costly External Finance, Corporate Investment, and the Subprime Mortgage Credit Crisis." Journal of Financial Economics 97: 418-35.

European Commission. 2005. "New and Updated Budgetary Sensitivities for the EU Budgetary Surveillance." http://ec.europa.eu/economy_finance/economic_ governance/sgp/pdf/budg_sensitivities_092005_v02_en.pdf.

--. 2012. "EU Balance of Payments Assistance for Latvia: Foundations of Success." Occasional Paper no. 120 (November).

Greenlaw, David, Jan Hatzius, Anil Kashyap, and Hyun Song Shin. 2008. "Leveraged Losses: Lessons from the Mortgage Market Meltdown." U.S. Monetary Policy Forum Report No. 2, Rosenberg Institute, Brandeis International Business School and Initiative on Global Markets, and the University of Chicago Graduate School of Business.

Hazans, Mihails. 2007. "Coping with Growth and Emigration: Latvian Labor Market before and after EU Accession." http://ssm.com/abstract=971198.

--. 2011. "The Changing Face of Latvian Emigration, 2000-2010." In Latvia-- Human Development Report 2010/2011: National Identity, Mobility and Capability, ed. by Brigita Zepa and Evija KJave. Riga: Advanced Social and Political Research Institute of the University of Latvia.

--2013. "Emigration from Latvia: Recent Trends and Economic Impact." In Coping with Emigration: Policy Challenges and Responses in Baltic and East European Countries. Paris: OECD Publications.

--, and Kaia Phillips. 2010. "The Post-Enlargement Migration Experience in the Baltic Labor Markets." In EU Labor Markets after Post-Enlargement Migration, ed. by Martin Kahanec and Klaus F. Zimmermann. Berlin-Heidelberg: Springer.

IMF. 2009. "Republic of Latvia: Request for Stand-By Arrangement." Country Report no. 09/3. Washington, D.C.: International Monetary Fund.

--. 2010. "Republic of Latvia: 2010 Article IV Consultation." Country Report no. 10/356 (December). Washington, D.C.: International Monetary Fund.

Kang, Joong Shik, and Jay Shambaugh. 2013. "Many Paths to the Same End Point." Working Paper no. 13/69. Washington, D.C.: International Monetary Fund.

Krasnopjorovs, Oleg. 2013. "Krugman's Latvian Adventures." Blog dated September 24 on Makroekonomika.lv (site maintained by the Bank of Latvia) at http://www.macroeconomics.lv/krugmans-latvian-adventures.

Lown, Cara, and Donald Morgan. 2006. "The Credit Cycle and the Business Cycle: New Findings Using the Loan Officer Opinion Survey." Journal of Money, Credit and Banking 38: 1575-97.

Magud, Nicolas, Carmen Reinhart, and Esteban Vesperoni. 2012. "Capital Inflows, Exchange Rate Flexibility, and Credit Booms." Working Paper no. 12/41. Washington, D.C.: International Monetary Fund.

Mejihovs, Aleksejs, and Anna Zasova. 2009. "Assessment of the Natural Rate of Unemployment and Capacity Utilisation in Latvia." Baltic Journal of Economics 9, no. 2: 25-46.

Mody, Ashoka. 2013. "Latvia in the Eurozone: A Bet with No Upside." Improving Economic Policy (blog), Bruegel.org, dated June 13.

Peek, Joe, and Eric Rosengren. 2000. "Collateral Damage: Effects of the Japanese Real Estate Collapse on Credit Availability and Real Activity in the United States." American Economic Review 90: pp. 30-45.

Purfield, Catriona, and Christoph Rosenberg. 2010. "Adjustment under a Currency Peg: Estonia, Latvia and Lithuania during the Global Financial Crisis 2008-09." Working Paper no. 10/213. Washington, D.C.: International Monetary Fund.

Reis, Ricardo. 2013. "The Portuguese Slump-Crash and the Euro-Crisis." Brookings Papers on Economic Activity, Spring.

Rimsevics, Ilmars. 2007. "Standing Firm against Manipulations with Lats' Exchange Rate." Diena, February 16.

--. 2009. "Stability of the Lats will be Maintained." Press release dated June 9, available on the Bank of Latvia website, http://www.bank.lv/en/for-media/ stability-of-the-lats/stability-of-the-lats-will-be-maintained.

Romer, Christina. 1990. "The Great Crash and the Onset of the Great Depression." Quarterly Journal of Economics 15, no. 3: 597-624.

Rosenberg, Christoph. 2009. "Why the IMF Supports the Latvian Currency Peg." Economonitor.com (blog), dated January 6.

--, and Marcel Tirpak. 2008. "Determinants of Foreign Currency Borrowing in the New Member States of the EU." Working Paper no. P/08/173. Washington, D.C.: International Monetary Fund.

Schadler, Susan, Ashoka Mody, Abdul Abiad, and Daniel Leigh. 2006. "Growth in the Central and Eastern European Countries of the European Union." IMF Occasional Paper no. 252. Washington, D.C.: International Monetary Fund.

Shambaugh, Jay, and Joong Shik Kang. 2013. "The Evolution of Current Account Deficits in the Euro Area Periphery and the Baltics; Many Paths to the Same Endpoint." Working Paper no. 13/169. Washington, D.C.: International Monetary Fund.

Staiger, Douglas, James H. Stock, and Mark W. Watson. 1996. "How Precise Are Estimates of the Natural Rate of Unemployment?" NBER Working Paper no. 5477. Cambridge, Mass.: National Bureau of Economic Research.

Vamvakidis, Athanasios. 2008. "Potential Growth Estimates in Emerging Europe Based on a Growth Model." Box 8 in the EUR Regional Economic Outlook (April) Washington, D.C.: International Monetary Fund.

Vanags, Alf. 2013. "Latvia's Exports: The Real 'Success' Story." BICEPS Research Report. http://biceps.org/assets/docs/izpetes-zinojumi/Latvias_exports.

Comments and Discussion

COMMENT BY

KRISTIN J. FORBES This paper by Olivier Blanchard, Mark Griffiths, and Bertrand Gruss offers an excellent and accessible description of the macroeconomic boom-and-bust cycle in Latvia from 2000 to 2012. It presents the key facts and data as well as is possible given the serious data challenges. It provides an unbiased analysis of different explanations for the country's crisis and recovery, which is much appreciated, given how heated macroeconomists' discussions on what actually happened in Latvia often have been.

Nonetheless, many readers of this volume may still be wondering, "Why Latvia?" Why should this distinguished group of authors and scholars dedicate an entire session to understanding the details of a small country that operated under a number of unique circumstances--especially during a period when many larger and more systemic economies were under severe economic distress? I will begin by answering this question and explaining why Latvia's experience is worth closer investigation. I will then discuss three major insights from the paper--as well as the follow-up questions they inspire.

WHY STUDY LATVIA? Latvia provides one of the few examples of a country that did not follow the standard recommended response to a balance-of-payments crisis: allow a rapid devaluation of the currency in order to regain competitiveness and reduce the need for external financing. Instead, Latvia chose to maintain its currency peg and attempt an "internal devaluation," that is, improve competitiveness through reductions in relative real wages. This process involved a period of sharply higher interest rates, a severe recession, and a fiscal contraction. In the past, other countries faced with similar balance-of-payments crises often hoped to adopt this strategy and avoid a currency devaluation, but most countries that embarked on this strategy soon abandoned it. Latvia is one of the few examples of a country that persisted on this difficult path, avoided a currency devaluation, and accomplished an internal devaluation--much to the surprise of many economists and much faster than anyone expected.

[FIGURE 1 OMITTED]

In order to understand exactly how Latvia's approach differs from the standard crisis response, it is useful to revisit a basic open-economy macroeconomics model. I will use what any graduate from MIT's Ph.D. or M.B.A. program calls the BB-NN model (which is basically a variant of the Salter-Swan or dependent economy model). In this model, a country attempts to meet three constraints. First is the "external constraint," represented in my figure 1 by the BB line, which requires that a country is in balance-of-payments equilibrium. Second is the "internal constraint," represented by the AW line, which requires that a country has output at potential. Third is the rather nebulous "social peace" line, represented by the P line, which requires that real wages are above a level w in order to avoid protests and riots. The lines are graphed relative to output (Y) on the horizontal axis and competitiveness (the ratio of the exchange rate to wages, or e/w) on the vertical axis.

In many examples, especially after a period of rapid growth financed by large capital inflows, a country finds itself in a situation like that shown in figure 1. The three lines in the model do not all intersect at any one point and the country is in the position marked by the dot A, labeled "Latvia 2006." The country has output Y above potential and is to the right of the NN line. The country also has a large balance-of-payments deficit and is to the right of the BB line. Real wages are still high enough (helped by the overvalued exchange rate) that workers are "at peace" and not protesting, so the country still rests on the P line.

However, a country in such a situation is not able to stay at point A. At some point foreigners are no longer willing to finance the current account deficit and the country does not have sufficient reserves to maintain the overvalued exchange rate. The most common response to this balance-of-payments crisis is for a country to move to point B, the "Devaluation Option." This involves a major and quick currency depreciation (e[up arrow]) and a sharp recession (Y[down arrow]). The country quickly moves to external balance, but the decline in real wages often causes protests and riots, the lack of social peace represented here by the country being located above the P line.

Latvia chose not to follow this standard response to a balance-of-payments crisis. Instead of devaluing its currency, it chose to keep e constant and instead move to a point such as C (labeled "Latvia 2011"). This involved a more gradual decline in real wages (w[down arrow]), which moved the country up on the graph, but not as far as if the currency had been devalued. It also involved a sharp increase in interest rates and a recession (Y[down arrow]), which may have moved the country more or less to the left on the graph relative to what would have occurred with a devaluation. (We will return to this question later.) The response also involved outward emigration, which shifted the NN line to the left.

The key point from figure 1 is that Latvia chose a different alternative than most other countries choose when responding to crises. To further illustrate how unique the Latvian response was, my figure 2 shows how countries responded to the period of global financial turmoil from 2007-11 (based on a sample of 85 countries). The graph basically shows the number of countries that chose to (i) sharply increase interest rates; (ii) allow large currency depreciations; (iii) intervene in foreign exchange markets using large amounts of reserves; and (iv) increase controls on capital outflows (details are available in Forbes and Klein 2013). The graph shows that the most popular response was to allow currency depreciations--which Latvia avoided. Only a few countries elected to raise interest rates--especially from 2009 onwards.

[FIGURE 2 OMITTED]

Returning to the case of Latvia, many economists did not think that raising interest rates sharply, undergoing a severe recession, maintaining a currency peg, and undergoing an internal devaluation instead of a currency depreciation would work. Many believed Latvia would not have the political will to persist with these difficult macroeconomic measures over the long period required to complete the adjustment. Others worried that given nominal wage rigidities, it would be impossible to generate the needed decline in real wages and improvement in competitiveness without a currency adjustment. Latvia proved the skeptics wrong. Several years after the crisis began, it is now safe to say that Latvia provides a model of an alternative response to a balance-of-payments crisis for countries that do not wish to devalue their currencies. Therefore, understanding Latvia's experience and how this adjustment occurred is important and well worth a case study.

Not only does Latvia provide a key example of a country that adopted an alternative response to a balance-of-payments crisis, it also provides insights on two other key debates in international macroeconomics: the advantages and disadvantages of large, front-loaded fiscal consolidation and of free capital mobility. Latvia is often cited by commentators on each side of these major debates as an example for or against a specific policy. The paper by Blanchard, Griffiths, and Grass is therefore extremely useful to better understand exactly what occurred in Latvia and how the country's experience is being used (or abused) to make specific points.

[FIGURE 3 OMITTED]

Here is an example. One prominent debate occurring in the blogosphere is whether the Latvian strategy should be a model for other countries. One side of this debate is expressed by Paul Krugman (2012), who asks, "Would you have expected that Latvia be lionized as the hero of the crisis?" To make his point, he graphs real GDP for Latvia and several other small countries in the region with real GDP indexed to 100 at its peak for each country. Krugman's graph is recreated here (using the identical data) as my figure 3. It shows that Latvia has clearly performed worse than the others.

On the other side of this debate, however, are Benn Steil and Dinah Walker of the Geoeconomics Center at the Council on Foreign Relations. They argue that one should consider the trends in GDP when GDP is indexed at 100 from the post-2007 trough rather than from the peak. As shown in my figure 4 below (which is copied from their website), Latvia now performs second best in the group.

Resolving this debate (and the many others) in the blogosphere that involve Latvia is beyond the scope of my comments today. But the key point is that the paper by Blanchard, Griffiths, and Gruss is an extremely useful reference to understand the different lines drawn in these discussions. For example, their paper documents the substantial boom and overheating in Latvia that preceded its crisis--and which has led to the differing interpretations of Latvia's GDP path as a model of above-average or below-average performance.

[FIGURE 4 OMITTED]

KEY INSIGHTS This paper also provides a number of useful insights and lessons. While it is impossible to generalize from events in one small and in many ways unique country to the rest of the world--and the authors are careful not to do so--Latvia's experience provides useful evidence to support three issues.

First, the crisis in Latvia was in many ways a standard balance-of-payments crisis. It was preceded by all of the typical vulnerabilities that generally precede crises, including large capital inflows from abroad, negative real interest rates, a bubble in housing, a sharp increase in imports, imports dominated by consumption goods rather than investment, rapid growth in private credit, large and unhedged borrowing in foreign currency, a very large current account deficit, and foreign capital inflows predominately in the form of "hot" money instead of foreign direct investment. (See Frankel and Saravelos 2012.) Standard early-warning models were identifying Latvia as being highly vulnerable to a standard balance-of-payments crisis.

Despite all the warning lights blinking bright red, the government was unable to prevent the economy from overheating and avoid a severe crisis. Several steps were taken to reduce credit growth, but these were too little, too late. This example suggests that even when a crisis is relatively easy to identify in advance, it is still extremely difficult for policymakers to take preventive measures in a timely fashion. Is there a way to develop more automatic stabilizers to prevent crises if policymakers cannot act in time due to political or other constraints? Unfortunately, the evidence in the paper further indicates that any type of automatic stabilizers would be difficult to construct. For example, figure 5 in the paper shows the real-time, cyclically adjusted fiscal balance in Latvia and how this number was substantially revised over time. This challenge to access accurate information on key statistics in a timely fashion highlights the challenge in preventing bubbles and crises in advance, even through automatic adjustment.

Second, the paper highlights the importance of quickly adjusting competitiveness. A key factor in Latvia's ability to persist with its alternative strategy was its surprisingly rapid internal devaluation. This appears to have occurred due to a combination of productivity increases, layoffs, and emigration. Its success, however, raises even more questions on exactly how it was accomplished and if the experience could be replicated. For example, how important was Latvia's small size to its ability to boost exports as part of its recovery? Does this experience suggest that labor market flexibility (firing) and mobility (emigration) are the key criteria to regain competitiveness with a fixed currency? If so, should countries hoping to share a currency (such as the euro) focus on these criteria rather than sharing fiscal risks (which is often included equally with labor market flexibility and mobility as a key criterion for an optimal currency area)? What are the long-term repercussions of this type of internal devaluation-- especially if it involves policies such as the emigration of skilled workers?

Perhaps the most important question raised by Latvia's example concerns the size of the contraction. If a country chooses to respond to a balance-of-payments crisis through an internal devaluation instead of a currency devaluation, will this generate a larger or smaller recession? Returning to my figure 1, will the country's leftward movement (decline in Y) from point A to point B be greater or less than its movement to point C? It is impossible to know the counterfactual, especially given the global turmoil that immediately followed Latvia's bust, but better understanding why the contraction may have been less (or greater) under Latvia's crisis-response strategy would be a useful addition to the paper.

[FIGURE 5 OMITTED]

To further make this point, consider the comparison shown in my figure 5. The graphs show the effect on real GDP over the six quarters following a country's decision to increase interest rates sharply or allow a major currency depreciation (both in quarter 0) during the global financial crisis (2007-11). The effects are estimated using a propensity-score matching methodology to construct the counterfactuals of what would have happened to GDP growth in each country if it had not had the increase in interest rates or currency depreciation. The panel on the left shows that countries that raised interest rates substantially--a key tenet of Latvia's strategy--generally saw a sharp and significant decline in real GDP growth immediately and over the next three quarters relative to the counterfactual. The panel on the right shows that countries that allowed large currency depreciations initially saw a contraction in GDP (over the first quarter) but then experienced an improvement in real GDP after several quarters, so that real GDP growth was significantly greater than the counterfactual after six quarters. Could Latvia have experienced this type of more rapid recovery if it had followed the standard crisis response and allowed its currency to depreciate?

A final key insight (and corresponding question) from this paper concerns the importance of supporting banks in order to enable a country to recover from a crisis. Latvia's banking system managed the crisis better than that in many other countries. One key factor highlighted in the paper was the support of foreign parent banks for local subsidiaries. Another key factor was that because the currency did not depreciate, banks were not faced with as many bankruptcies and the corresponding increase in nonperforming loans, which generally follow when companies and consumers have unhedged liabilities (even if the banks had directly hedged their assets and liabilities against currency risk). Moreover, although one major domestic bank in Latvia required sovereign support, this did not create the large fiscal burden that occurred in other countries, such as Ireland.

The relative success in stabilizing Latvia's banking system was undoubtedly a key component of the country's overall recovery. But returning to the question of why the Latvian experience is worth closer investigation--how did Latvia's decision not to devalue influence the recovery of its banking system? If Latvia had chosen the standard currency-devaluation response, would this have generated a widespread increase in non-performing loans that would have caused more widespread banking collapses? Could this strategy have risked the support of foreign parents (or generated more financial support from them)? A closer look at how Latvia's decision not to devalue its currency interacted with the financial system would add another useful dimension to the paper and help one to better understand one potential benefit of this alternative crisis response.

To conclude, this paper is an excellent case study of what happened in Latvia over the 2000s. It illustrates a textbook case of the standard buildup to a balance-of-payments crisis and provides a rare example of a nonstandard response. The case study is a superb example for teaching concepts and for better understanding one of the options that countries consider when faced with a balance-of-payments crisis. Latvia's experience provides useful insights into several major questions, such as why it is so difficult for policymakers to prick bubbles, how a country can regain competiveness with a fixed exchange rate, and the importance of a strong banking system. Each insight, however, generates more questions--especially the question of how Latvia's example could be generalized to other countries. More case studies in this vein--of large and small countries alike--would be helpful to better understand the policy options countries face.

REFERENCES FOR THE FORBES COMMENT

Forbes, Kristin, and Michael Klein. 2013. "Pick Your Poison: The Choices and Consequences of Policy Responses to Crises." Working Paper no. 5062-13. Cambridge, Mass.: MIT Sloan School of Management.

Frankel, Jeffrey, and George Saravelos. 2012. "Can Leading Indicators Assess Country Vulnerability? Evidence from the 2008-09 Global Financial Crisis." Journal of International Economics 87 (2): 216-31.

Krugman, Paul. 2012. "Peripheral Performance." Conscience of a Liberal (blog), New York Times, June 14. http://krugman.blogs.nytimes.com/2012/06/14/ peripheral-performance/.

Steil, Benn, and Dinah Walker. 2012. "Iceland's Post-Crisis 'Miracle' Revisited." Geo-Graphics (blog), July 2, Center for Geoeconomic Studies, Council on Foreign Relations (http://blogs.cfr.org/geographics/2012/07/02/postcrisis/).

COMMENT BY

PAUL KRUGMAN Olivier Blanchard, Mark Griffiths, and Bertrand Gruss have given us a terrific paper on Latvia, welcome for its tone as well as its content. Latvia has become a symbol in the fiscal policy wars, with austerity advocates elevating it to iconic status; the temptation must have been strong either to validate that elevation or to turn the paper into an exercise in debunkery. Instead, the authors give us a detailed, balanced account--one that highlights just how odd, how inconsistent with the orthodoxies of either side, the Latvian experience seems to be.

Let me dive right into the two big issues the paper raises: the puzzle of Latvia's output gap and the puzzle of its internal devaluation.

Here is what is known: Latvia suffered a huge, Depression-level economic contraction after 2007, followed eventually by a fast but as yet incomplete bounce-back--which the latest data suggest may be slowing--that has left unemployment much higher than it was pre-crisis. Actually, Latvia's numbers from 2007 to 2013 look fairly similar to those for the United States from 1929 to 1935. Today, everyone considers America in 1935 to have been still in the depths of the Great Depression, so if one looks at Latvia through the same lens the country doesn't look very good--better than Greece, perhaps, but not good.

However, the Latvian authorities tell a very different story, and the authors basically agree. The authors argue that 2007 is a misleading base, that the Latvian economy on the eve of the crisis was wildly overheated, with a positive output gap of something like 12 percent. And they correspondingly conclude that Latvia has in large part already recovered more or less fully. They do not arrive at this conclusion lightly. But I do think it is worth asking how plausible it is.

First of all, on a conceptual level, how does an economy get to operate far above capacity? We all understand operating below capacity: producers may fail to produce as much as they would prefer to if there is not enough demand for their products. But how does excess demand induce producers to produce more than they want to? New Keynesian models actually have something of an answer: The economy is monopolistically competitive, so producers in general charge prices above marginal cost and are hence willing to produce more given the demand. But there has to be some limit to this margin. Is 12 percent really plausible?

Second, how often does one see the kind of huge positive gap that is being posited for Latvia? Or to ask a question that can actually be answered, how often does the IMF estimate output gaps that big? I have gone through the IMF's World Economic Outlook Database, looking at all advanced countries since 1980, to identify double-digit positive output gaps. Here's the full list:

Estonia 2007
Greece 2007
Italy 1980
Luxembourg 1991

I have no idea what was going on in Italy in 1980 or in Luxembourg in 1991. I doubt that anyone believes that Greece was operating at 10 percent above capacity in 2007. Surely what is showing up here is a problem with the methods the IMF uses to estimate potential output. Basically, the Fund uses a weighted average of actual output over time. This automatically interprets any sustained decline in actual output as a decline in potential, and it causes that re-estimate to propagate backward through time. So the catastrophe in Greece ends up leading to the basically silly conclusion that there was a hugely overheated economy there before the crisis.

In addition, whatever is going on in Estonia presumably bears some relationship to what is going on in Latvia. The point is, if Latvia really was as hugely over capacity as the authors claim--and to be fair, they do not use the filtering method but instead make a careful assessment of unemployment and inflation--it represents a more or less unique case.

And arguing that Latvia was vastly over capacity in 2007 has another, perhaps surprising implication: It makes much of the debate over both austerity and internal devaluation moot.

On austerity: If one were really looking at an economy with a double-digit inflationary output gap, even the most ultra-Keynesian Keynesian would call for fiscal austerity. Once one grants the output gap interpretation, the fiscal policy debate evaporates. To an important degree, the internal-versus-external devaluation debate evaporates as well.

My figure 1 plots Latvian growth, at an annual rate, versus the one-year change in the current account balance as a percentage of GDP. There was a strong relationship both before and after the crisis, but if anything it was stronger before the crisis. Given the slump in Latvian output after 2007, one should have expected a huge external adjustment from that fact alone. Maybe Latvia did not need a devaluation of any kind, external or internal. Maybe it was not overvalued, just overheated.

[FIGURE 1 OMITTED]

That said, the authors also provide evidence of a substantial internal devaluation, at least as measured by unit labor costs. Oddly, however, almost none of this comes by means of lower wages: wages in manufacturing have been every bit as flat as those of us who warned about downward nominal wage rigidity would have predicted. Instead, what one sees is a rapid rise in productivity, which the authors suggest is the result of eliminating X-inefficiency.

However, there may also be an alternative interpretation. Latvia is a relatively poor European country playing catch-up, and it had rapid productivity growth before as well as after the crisis. My figure 2 shows aggregate labor productivity from Eurostat. Maybe Latvia just had an impressive productivity trend owing to its particular position in the European system and simply returned to that trend after a brief setback.

How does that bear on the internal devaluation debate? If Latvia had very high productivity growth, for whatever reason, the serious thing advocates of currency flexibility worry about, namely the downward rigidity of nominal wages, was not a binding constraint.

Suppose, instead, this story: Latvia was a hugely, perhaps uniquely overheated economy that even a Keynesian would agree needed a lot of fiscal austerity, a country with very high rates of productivity growth making wage stickiness irrelevant. I am not sure I believe this story, but for those who do, what lessons does Latvia hold for other countries and for the eurozone in general?

[FIGURE 2 OMITTED]

The answer, in brief, is none. Latvia's story as I have just told it looks nothing like anything we have seen in the past, and probably not like anything we are likely to see in the future--not even in Latvia's future. So I am a little puzzled by the authors' sanguine view about Latvia's entry into the euro. The next time there is a euro crisis--and there will be another one, someday--there is no reason to believe that anyone will be able to adjust in the way that Latvia, just possibly, has.

GENERAL DISCUSSION Steven Davis was struck by the suggestion in the authors' paper that Latvia contained prior inefficiencies that were removed under the pressure of the financial crisis. He could not think of another country with that experience, although some industries came to mind, including the iron ore industry in North America and Europe, which made tremendous improvements in productivity without any technological revolution after worldwide steel demand collapsed in the early 1980s, that is, entirely through changes in work practices. Davis felt that achieving similar gains at the scale of even a small country was more impressive, and it would be useful to better understand how those gains came about.

Susanto Basu was impressed as well by the rise in Latvia's productivity, even from the very beginning of the crisis, whereas fixed costs, utilization, and other conventional measures would normally show a decline. He inquired whether it may have been due not to changes in sectoral composition but to changes in the composition of workers, since workers with lower productivity are normally the first ones to be fired. In a deep downturn, only the most productive workers would be retained, so perhaps labor productivity in Latvia wasn't growing as fast as it appeared to be and that wages were therefore falling much faster than they appeared to be. All of this would mean that internal devaluation was even more successful than one had thought.

John Haltiwanger asked whether compositional effects within sectors were the determinative factor in productivity increases, given that the authors showed sectoral composition was not determinative. Substantial productivity increases can be attained from such changing composition, particularly in countries undergoing reforms. Haltiwanger suggested that if Basu were correct about worker retention, in Latvia such within-firm productivity increases should be evident. Alternatively, Latvia's productivity growth could have stemmed from the reallocation of labor, as has happened in countries undergoing market reform, including China, India, and much of eastern Europe. What remained puzzling about Latvia was how quickly all of this happened.

David Romer pointed out that no internal devaluation (in the sense of a fall in prices relative to those of other countries) appeared to have actually happened in Latvia, so the question remains, What caused the export boom? He suggested three possible answers. One, implicit in Paul Krugman's comment, is that there might have been no export boom in the first place, only a natural response to what was happening to output and the cooling off of the economy. A second answer, which Olivier Blanchard had implicitly suggested in his presentation, was that because Latvian producers were earning such enormous profit margins they had incentives to drum up new customers. This answer appears to require that they were able to do so without much labor input, since productivity did not fall. It would also imply that Latvia's producers did not take advantage of that situation until their labor costs went down. Romer observed that the paper did not present evidence in favor of these implications of this explanation. A third explanation might simply be that "they got lucky," that is, that the export boom had nothing to do with the fall in unit labor cost, since that did not translate into lower prices.

Caroline Hoxby suggested that Latvia's export boom might be explained as stemming from the reallocation of labor between firms, returning to a possibility Haltiwanger had raised. She wondered whether the firms that gained a lot of the labor market share within their industry were the very ones that were already good exporters. Firms' propensity to export varies widely even in the same industry, so efficient firms could have taken business from inefficient ones. It also seemed to Hoxby that the productivity gains were indeed real, since the workers who relocated must have been hired to do work in Europe that was at least as productive as what they were previously doing in Latvia. She would like to know whether those gains would have happened in Latvia without the sharpness of the fiscal consolidation and the government's willingness to stick to the peg. The most important question for her was whether there was a causal link between those decisions and productivity or whether it was only a case of coincidence.

Justin Wolfers took issue with Hoxby's suggestion that the productivity gains were real, noting that growth by 50 percentage points in four years would normally suggest something was wrong with the data. The simplest way to test this would be to ask what the natural implications in the economy would be following such a sharp productivity boom and then look for them. Above all, one would expect to see a change in prices, yet in Latvia no major price changes were observed. This leaves open the question of what has actually been going on in the country.

Richard Cooper agreed with Romer that the authors' data showed no evidence of an internal devaluation in Latvia, and agreed with several others as well that the biggest puzzle was the rise in productivity. Internal devaluations, he added, can indeed be relevant to export booms, as in the case of Germany, which stimulated exports in 2007 by reducing labor taxes and raising the VAT, but such an internal devaluation did not occur in Latvia.

Benjamin Friedman raised the issue of Latvia's massive emigration. Half a decade in which 1.0 to 1.5 percent of the country's population emigrated each year would be comparable to 20 million people exiting the United States during the same period. Should we emphasize the fact that this population movement was mostly within Europe, and therefore think of it as similar to a migration of workers from Michigan to Arizona, that is, a natural movement of labor that did not disturb the country as a whole? Or should we instead think of it as similar to labor leaving the United States altogether? In the first case, unemployment would be much less of a problem if workers could relocate so easily, although as a "solution" that would risk igniting serious social unrest. Friedman also thought the matter raises the question of what determines whether somebody who has lost a job decides to exit the country. Krugman had suggested that emigration is a big part of the Latvia story, and not as a problem, because Latvia is actually more like Michigan in the analogy. If that is correct, the lesson may be that one cannot apply the Latvian experience to a "normal" country.

Paul Krugman added that in the European context, issues of emigration require one to consider the lack of fiscal integration there. By contrast, when people leave Michigan for Arizona, this does not create a problem of who is going to pay their social security and Medicare benefits. When Portuguese leave Portugal, however, there is a very real concern over who will be left to pay for the benefits needed by those people who remain behind in the old country.

Robert Gordon was struck by the authors' finding that Latvian productivity had risen for 11 years by about 8 percent per year, in striking contrast with the euro zone, where it barely rose at all between 1995 and 2011. As an emerging country rapidly catching up to western methods of production, Latvia had the latitude to make radical changes in its unit labor cost, something that the problem countries of Europe could not do. Italy has experienced zero productivity growth for the last decade, and the potential for cutting unit labor cost there seemed to Gordon vastly different than in Latvia. Even in the United States, although productivity briefly boomed during the recession, it has almost completely stagnated since.

Valerie Ramey was most surprised by the authors' finding that the natural rate of unemployment in Latvia is around 12 percent, an exceptionally high rate for a country with a low minimum wage, weak unions, and limited unemployment insurance and employment protection. To her, this strained credibility. Was there a gross mismeasurement of the price level and hence inflation, or a gross mismeasurement of the unemployment rate? Or did this show that something is fundamentally wrong about the Phillips curve itself?

Gregory Mankiw too wondered whether Latvian measures were to be trusted at face value. He questioned whether the measures of unemployment there are reliable, for example whether cultural differences may cause people to answer surveys about looking for work differently than they do elsewhere. The fuzziness of the concepts of unemployment and labor force could also be contributing to the surprising findings.

Christopher Carroll expanded on a point made by Gordon regarding Latvia's uniqueness, specifically that its economy had been growing so quickly for a long time leading up to the crisis. If the macroeconomic findings on the importance of habit formation published in the last dozen years are correct, it will be much easier for Latvia to readjust back downward to where the country was three years ago than it will be for Portugal to readjust to below where it was three years ago. He added that many of the stories about the relative ease or difficulty countries have had in adjusting strongly suggest that habits might matter for governments as much as they do for people.

Olivier Blanchard responded to the various points raised. Clearly a central question is what stands behind high productivity growth. Strong labor shedding, induced in part by the credit crunch, was surely a factor at the start. But now, many sectors have higher employment and higher output than before the crisis, so more than labor shedding is at work. Regarding composition effects, he noted that while there was no evidence of sectoral composition effects, there was some evidence of skill composition effects. At the same time, the proportion of unskilled workers in unemployment is higher than it was before the crisis. Regarding wage adjustments, he noted that while public sector wages were lowered by 20 percent, private sector wages dropped only slightly.

Blanchard had this to say regarding the comparison between internal and external devaluations: One might argue that an external devaluation would have decreased the pressure to achieve productivity improvements, and that it may have triggered sharp adverse balance sheet effects, in response to the sudden increase in the domestic value of foreign currency liabilities. As far as the small decrease in export prices and the behavior of exports, he noted that if exporters are price takers in foreign markets, and the effect of a decrease in unit labor costs shows up in profit margins rather than in prices, the increase in profit margins may lead new firms to enter the export market and existing firms to increase production. Given that Latvia is a small country and has a small market share, it may indeed be a price taker in most export markets, and this may be the explanation behind rapid export growth.

Finally, Blanchard very much agreed with Krugman that, in the context of sharp export growth before the crisis, the return to rapid export growth may not be that much of a surprise. Clearly, things would be and are very different in countries that stagnated before the crisis.

OLIVIER J. BLANCHARD

International Monetary Fund

MARK GRIFFITHS

International Monetary Fund

BERTRAND GRUSS

International Monetary Fund

(1.) From Dombrovskis' interview with CNBC at the EU summit in Brussels, quoted in "Krugman Can't Admit He Was Wrong on Austerity: Latvia PM" and published on the CNBC.com website at http://www.cnbc.com/id/100558455.

(2.) From Kingman's New York Times blog, "The Conscience of a Liberal," dated January 2, 2013, and entitled "Latvia, Once Again."

(3.) A lot has been written on Latvia. Two important references, which look at the various aspects of the boom and the bust, and from which we have benefited, are by Aslund and Dombrovskis (2011) and the set of articles in European Commission (2012).

(4.) See the online appendix to this paper for details of the two specifications, at http:// www.brookings.edu/about/projects/bpea.

(5.) In current prices, the ratio of consumption to GDP remained nearly constant, while the ratio of investment to GDP increased from 23 percent to 40 percent, reflecting a large increase in the relative price of investment goods.

(6.) We measure domestic demand here as private final consumption plus government final consumption plus gross capital formation.

(7.) This behavior is consistent with the findings of Magud, Reinhart, and Vesperoni (2012), who find that bank credit grows more rapidly in response to capital inflows in countries with less flexible exchange rate regimes, with a shift in composition toward foreign currency lending.

(8.) Throughout this paper, the reported ULCs for the whole economy or for individual sectors are constructed as the ratio of aggregate nominal compensation of employees over real gross value added (GVA) in the respective category. This ensures consistency when ULCs for the whole economy are compared with ULCs for individual sectors, since GDP is not available at the sector level.

(9.) Eurostat official data on house prices start only in 2006; even so, from 2006Q1 until the peak in 2008Q1 they show a 75 percent increase.

(10.) See online appendix for details. The cyclically adjusted fiscal balance is constructed as the headline balance minus 0.3 times the output gap, following a method proposed by the OECD and estimates of elasticity for Latvia from the European Commission (2005). The different vintages of output gap series are those estimated by the European Commission (in its autumn economic forecasts).

(11.) Ending the practice of allowing borrowers to state fictitious incomes and imposing maximum loan-to-value ratios did help stop the demand for real estate, nevertheless. And when prices started to fall, demand dried up, anticipating future price declines.

(12.) "The Latvian government didn't do much to stop this economic transformation. If anything, it stepped on the gas. Riga's new deputy mayor and millionaire Ainars Slesers, who served in the Latvian Parliament during the boom years, coined a phrase that is sure to become a symbol of the prevailing government attitude at the time: gazi grida (pedal to the metal)." From "Latvia's Tiger Economy Loses Its Bite," a journalistic report by Kristina Rizga published online by the Pulitzer Center on Crisis Reporting, http://pulitzercenter.org/articles/ latvias-tiger-economy-loses-its-bite.

(13.) On Portugal, see Blanchard (2007) and Reis (2013).

(14.) A first pass at the effects of the crisis on emerging market economies in general was presented in an earlier Brookings Paper (Blanchard, Das, and Faruqee 2010).

(15.) The decrease in imports seems large as a proportion of GDP, but the relevant denominator is domestic demand. The decrease in imports is equal to 60 percent of the decrease in domestic demand, roughly equal to the ratio of imports to domestic demand in 2008:1.

(16.) See online appendix. The estimated elasticity with respect to partner country GDP is 1.58; the estimated elasticity with respect to the real exchange rate is-0.21.

(17.) Latvian subsidiaries of Nordic banks include Swedbank, SEB, DNB, Danske Bank, and Nordea (the last two operate as branches rather than as subsidiaries).

(18.) Nonresident deposits in Latvia were mostly by CIS corporations, due to ease of transaction, geographical proximity, and language.

(19.) Calomiris and Mason (2003) focus on the Great Depression; Peek and Rosengren (2000) on Japan; Greenlaw and others (2008) and Duchin, Ozbas, and Sensoy (2010) on this crisis; and Claessens, Kose, and Terrenes (2009) on a number of crunches.

(20.) It is not clear whether the right parameter one should look at is the response of output growth to credit growth, or instead the response of output growth to the change in the credit-to-GDP ratio. The studies' results that we mention are stated in terms of the first.

(21.) It is not clear that credit growth is the right metric to relate to demand growth. As one might expect, the decrease in new loans was much larger. After peaking in 2006Q4, the flow of new loans (constructed from the change in the credit stock and an estimated amortization flow based on the maturity structure of bank loans) had decreased by 65 percent in real terms by end-2008 and by 85 percent by end-2009. A related metric, constructed as the difference in the flow of credit relative to GDP (see Biggs, Mayer, and Pick 2009) would suggest that the contribution of credit to demand growth, or credit impulse, was-19 percent of GDP over the period 2008Q3 to 2009Q3.

(22.) For comparison, in her study of the effects of uncertainty on spending at the onset of the Great Depression, Christina Romer (1990) found that car registrations declined by 24 percent from September 1929 to January 1930.

(23.) Anecdotally, Latvia, which does not produce cars, became a car exporter for a few months in late 2009 and early 2010, as dealers, unable to sell their inventory of cars at home, sold them to foreign dealers.

(24.) Although all the revenue measures, adding to 2.5 percent of GDP, were introduced, only about 1.5 percent of GDP in expenditure cuts are estimated to have been implemented.

(25.) Public sector wages had increased only slightly more than private sector wages during the boom, but were 30 percent higher.

(26.) The cyclically adjusted fiscal balance in figure 9 is defined as the headline balance-to-GDP ratio minus the output gap times an overall budgetary sensitivity parameter, based on European Commission (2005) estimates, of 0.3.

(27.) One has to wonder about such large spreads for a country where the ratio of net debt to GDP was still around 10 percent (1200 basis points imply an annual payment of 12 percent of the nominal value of the debt).

(28.) See, for example, Krasnopjorovs (2013).

(29.) NACE is the European community's statistical system for classifying economic activity. Due to data availability, data at the one-digit level were used for (B) Mining and quarrying; (Q) Human health and social work activities; and (R) Arts, entertainment, and recreation.

(30.) See the next section and the online appendix for the results of estimation of Phillips curve relations. The time series is however too short to reach strong conclusions. A Phillips curve specification, allowing for an effect of public sector on private sector wage inflation, does not yield conclusive results. But it may be that there was a one-time strong effect in 2009.

(31.) In view of this fact, one of the editors has questioned whether the adjustment should be described as an "internal devaluation." We think that it should, since it reflects an adjustment in unit labor costs at a fixed exchange rate, which is the essence of an internal devaluation.

(32.) However, figure 12 seems to have an unexplained trend. From 2000 to 2007, export prices increased faster than the partner country price index. Strictly speaking, this is hard to reconcile with the behavior of export prices since 2007, which appears largely to reflect pricing to market. While one can think of a number of measurement problems (for example, a shift toward higher quality and thus higher priced goods) that could explain this fact, we do not have convincing evidence.

(33.) Again, the data present a minor puzzle. One might have expected the GDP deflator, which reflects both the price of tradables and the price of nontradables, to decline more than the export price deflator. This was not the case. As can be seen from figure 4, the GDP deflator declined less. Even though Latvia's domestic prices fell (its CPI at constant tax rates fell by almost 10 percent), the fall in world prices during the crisis was exceptional and even greater.

(34.) While both internal and external devaluations imply the same foreign-exchange-induced balance sheet effects, their timing can be different. The balance sheet effect is instantaneous in the case of a nominal exchange rate adjustment. But it happens over time in the case of an internal devaluation, and thus allows more time to adjust.

(35.) More recently, Latvijas Krajbanka, a medium-size bank, had an intervention in November 2011 after fraud was discovered. But the bank did not receive state aid and was liquidated in early 2012.

(36.) Interestingly, Eurostat estimates PPP GDP per capita to now be 9 percent above its 2008 peak.

(37.) Latvia's implementation of the Schengen agreement started in December 2007. although Ireland, the United Kingdom, and Sweden opened their labor markets to those from the new EU member states in 2004. Thus, part of the post-2007 increase may also reflect easier access to labor markets in continental Europe.

(38.) To a first approximation, emigration should have no effect on the long-run Beveridge curve. In the short run, however, it could lead to a shift inward as unemployment declines due to vacancies. FFowever, there is no evidence that as the emigration rate has returned to precrisis levels the Beveridge curve has shifted back out.

(39.) The puzzle is not unique to Latvia. The natural rate appears to be high in a number of central and eastern European countries. In the other two Baltic countries, the average unemployment rate since 2000 has been above 10 percent (10.3 percent in Estonia and 11.8 percent in Lithuania).

(40.) In related work, we have explored how trust in government depends on both the level and the change in the unemployment rate, as well as other controls. Using survey data for EU countries from the Eurobarometer, we have found roughly equal coefficients on the two variables. A decrease of one percentage point in the unemployment rate can offset a one percentage point higher level of the unemployment rate.

(41.) The data and their descriptions are available on a webpage maintained by the Central Statistical Bureau of the Republic of Latvia: http://www.csb.gov.lv/en/statistikas-temas/ metodologija/occupied-posts-and-job-vacancies-36334.html.

(42.) See the online appendix for alternative model specifications and alternative measures for consumer price inflation.