Inflation in the Baltics: on its way down?

Each month Eurostat arranges a new round of the unofficial competition in “consumer price inflation” as it publishes a new round of inflation figures for the 27 EU countries. For a while the three Baltic countries have dominated in the top (or bottom as you like it!) with inflation rates substantially above the levels seen in the old EU countries. Figure 1 below shows the year-on-year HICP consumer price in the Baltic countries since 1997. The price inflation is now in the double digit range for all three countries; the only other EU country with similarly inflation pressure is Bulgaria. This raises the simple question: what drives – and what has driven – inflation in the Baltic countries?

blogbalticbaltic_inflation_1_640.gif

Inflation in the Baltics is excessively high. The dream of early membership of the EMU has had to be shelved as the inflation rate in all three countries is way above the reference value of the Maastricht inflation criterion. The postponement of EMU membership has left a political vacuum and the public has become impatient as the promised accession date is repeatedly push into the future. The high inflation rates have also reached a point where they constitute problems by themselves. Nominal interest rates follow the Eurozone rate closely, so real interest rates are strongly negative, which may lead to misallocation of capital. There are also increasing complaints from groups in society who feel that they are falling behind as they see their living standard undermined by rising pricing. Inflation has moved from a “solved problem” to one of the most imminent problems on the economic policymaking agenda.

As with many other economic problems, it came all so suddenly. Estonia and Lithuania adopted currency boards and Latvia a close peg in the beginning of the 1990s to escape the hyperinflation raging in Russia. With the exchange rate as a “nominal anchor”, inflation gradually declined from three to one-digit levels during the period 1992-97. Inflation was relatively low and stable from 1999 until the middle of 2006, but then suddenly accelerated to reach levels above 10 percent per year in the beginning of 2008.

It is noticeable that GDP and price levels are almost always very closely correlated. Countries with low income levels have generally low prices – and vice versa. Figure 2 shows nominal GDP per capita in euro and the GDP deflator in 2006 for 26 EU countries. The correlation is very tight and it remains if, for example, the income measure is taken to be GDP per capita in purchasing power adjusted terms or the GDP deflator is replaced by the consumer price index. The tight correlation suggests a connection between the recent inflation upturn and the rapid economic growth experienced in the Baltics during the period 2005-07.

blogbalticbaltic_inflation_2_640.gif 

The trouble with the tight correlation between relative GDP and price levels is that it does not provide much information on the factors causing the correlation. The best candidate is the Balassa-Samuelson effect, which predicts higher consumer price inflation if productivity increases are higher in the traded sectors than in the non-traded sectors of the economy. High productivity growth in traded sectors leads to wage increases that spill over into non-traded sectors and consequently lead to higher price inflation in the non-traded sector. In some sense, structural inflation caused by the Balassa-Samuelson effect is a benign equilibrium phenomenon as it reflects high economic growth in (parts of) the economy.

Research has generally estimated the Balassa-Samuelson effect account for at most 1-2 percentage points of the annual inflation. These estimates may be relatively low given the substantial productivity increases in manufacturing in recent years. Also, some of the analyses fail to take into account that traded products often contain a large component of non-traded input when bought by consumers. The non-traded inputs comprise, for example, domestic transportation, distribution, marketing, sale, warranty repairs, service contracts etc.

Among other “equilibrium explanations” is an upward trend in the quality of both domestically produced and imported products. A few years ago, a shirt was a shirt; now a shirt is only a shirt if a fancy designer label is attached to it. If the statistics authorities fail to adjust fully for such “quality improvements”, the result would be higher measured inflation. It is also possible that firms with market power increase their mark-ups when the purchasing power increases, in which case the result would again be a positive link between income and price levels.

The discussion above suggests that the structural or equilibrium inflation rates in the Baltic countries are somewhat higher than what is typically seen in the old EU countries in Western Europe. It does not, however, explain the rather swift changes in the inflation seen in Figure 1. It is possible to point to a number of specific factors, including the cyclical position, that have been of importance.

First, the Russian crisis in August 1998 hit the Baltic countries very hard; the resulting economic downturns along with cheaper import from Russia led to lower inflation. Second, Lithuania’s shift from the dollar to the euro brought about an effective appreciation of the litas with negative inflation as a consequence. Third, the accession to the European Union led to markedly higher food prices as the Baltic countries became part of the internal market for agricultural products. The accession also meant increases on excise taxes on energy, tobacco and alcohol. It is noticeable that the price shocks after EU accession did not die out at once but set off higher inflation for a longer period of time. Fourth, in the middle of 2006 the inflation dynamics change abruptly and inflation accelerates from 4-6 percent per year to more than double this level. This acceleration coincides with rapid growth and a very high rate of capacity utilisation in all three countries. Finally, changes of energy prices, import prices and controlled prices have also played a role in some cases, but the effects are not evident from Figure 1 (and are also difficult to establish econometrically).

Overall, the developments of inflation in the Baltics can largely be explained by the “usual suspects” (shocks and capacity utilisation) although the trend level of inflation is likely to be higher than in the Eurozone because of rapid economic growth. The big question mark, however, concerns the autoregressive properties of the inflation processes. Some studies (see references below) suggest that there is a substantial forward-looking component in the inflation processes in the Baltic countries, but such findings are not very reliable because of correlation between the lagged and leaded inflation terms. A casual glance at Figure 1, however, suggests that inflation at different times have changed substantially in a short time; this may support the view that the Baltic inflation processes exhibit limited inertia.

Economic growth is undoubtedly on its way down and that should ease inflationary pressures as probably already witnessed during recent months in Estonia and Latvia. Energy prices have (hopefully!) topped and may fall over the next one or two years; food prices may follow the same pattern. The question is then whether the downward impact of these shocks will dominate over inertia carried over from the high inflation in 2006-07? In other words, will inflation rates return to trend levels around 3-5 percent or will backward-looking wage and price setting lead inflation to continue at recent high levels? My guess is that the inertia is limited and that a cooling of the Baltic economies will dampen inflation significantly. Hopefully, the cooling of the Baltic economies will not be too strong and/or long-lasting; Figure 2 illustrates all too well the income gap to the high income Western EU countries.

______________________

Masso, Jaan & Karsten Staehr (2005): “Inflation dynamics and nominal adjustment in the Baltic States”, Research in International Business and Finance, vol. 19, no. 2, pp. 281-303.

Dabusinskas, Aurelijus & Dmitry Kulikov (2007): “New Keynesian Phillips curve for Estonia, Latvia and Lithuania”, Working Papers of Eesti Pank, no. 7/2007, Bank of Estonia.

3 Responses to "Inflation in the Baltics: on its way down?"

  1. eparisi   April 30, 2008 at 4:39 pm

    I wonder how the Baltics’ GDP per capita performed in terms of purchasing power especially in the last decade. Was there some real catching-up during this period? Thanks for hints.

  2. Karsten Staehr   May 5, 2008 at 7:30 am

    To eparisi: According to Eurostat the purchasing power adjusted GDP per capita in the Baltic countries has increased from around 30% of the EU15 average in 1995 to around 50-60% in 2006. Thus, the Baltic countries have experienced a rapid catch-up in terms of international purchasing power within the last decade. (This is also clearly visible when one compares the consumption of cars, consumer electronics, foreign holidays, etc. in 1995 and 2006.) I have written a bit more about it in a piece from the Danish National Bank: http://www.nationalbanken.dk/C1256BE9004F6416/side/Monetary_Review_2007_4Quarter/$file/kap06.htm

  3. eparisi   May 6, 2008 at 12:03 pm

    Thanks!