Poland Business Cycle
This Country Focus analyses and interprets the statistical characteristics of the Polish business cycle. It also identifies leading and lagging variables and shows that the economic fluctuations in Poland differ to some extent from those in other emerging and mature economies, with Polish growth notably more volatile and government expenditure highly erratic. The available data on GDP growth suggest that the Polish economy is approaching the peak of the second business cycle since the start of economic transformation from a centrally planned to a market economy.
The current upswing is to some extent similar to the one of 1995-1997 which ended in large macroeconomic imbalances (increasing unemployment, spare capacity, widening fiscal and current account deficits). However, Poland now appears to be better positioned than after the last cycle and should be able to avoid a repeat of that outcome. Business cycles in emerging market economies Although the economic literature on business cycles is vast, only recently have some papers on business cycles in emerging market economies appeared.
Usually they analyse economic fluctuations within particular countries (e. g. Benczur and Ratfai, 2005) or make some cross-country comparisons (e. g. Aguayo et al. , 2004 or Carmignani, 2005). A common methodology used in the analysis of business cycles (based on observations of mature economies and economic theory) distinguishes pro-cyclical, counter-cyclical and a-cyclical variables. Pro-cyclical variables fluctuate together with GDP (e. g. industrial production, investment, employment, inflation), countercyclical variables against GDP (e. g. nemployment, net exports) and a-cyclical variables independently of GDP (e. g. real interest rates). 1 With respect to timing, the ‘stylised facts’ of the business cycle identify leading, lagging and coincident variables: leading variables move ahead of GDP (e. g. average labour productivity, inventory investment, money supply), lagging variables follow GDP (e. g. inflation, nominal interest rates) and coincident variables, as the name suggests, move coincidentally with GDP (e. g. industrial production, consumption, employment) (Snowdon & Vane, 2005, p. 306).
Overall, business cycles in emerging market economies (Carmignani, 2005)2 are not much different from those in mature economies (Snowdon and Vane, 2005, p. 306), By Michal Narozny* The business cycle in Poland: where do we stand? Highlights in this issue: • While on the whole not different from mature economies, the properties of the business cycle in Poland display some special characteristics • The current cycle seems to have reached a peak but the slowdown is likely not to be as pronounced as in the previous cycle Volume IV, Issue 9 03. 08. 2007 ECFIN COUNTRY FOCUS Directorate for the Economies of the Member States. The views expressed in the ECFIN Country Focus belong to the authors only and do not necessarily correspond to those of the Directorate-General for Economic and Financial Affairs or the European Commission.
Economic analysis from the European Commission’s Directorate-General for Economic and Financial Affairs Identification of direction and timing is key in business cycle analysis ECFIN Country Focus Volume IV, Issue 9 Page 2 but economies in transition (though they do not constitute a homogeneous group) display some specific characteristics: overall, the economy is much more volatile than in the euro area, which is the consequence of structural changes and catching-up, • shocks are slightly less persistent than in the euro area, and fluctuations consequently more frequent, • government consumption is more erratic than in the euro area, suggesting a significant discretionary element in fiscal policies, but not one that is necessarily aimed at cyclical stabilisation, • employment is a-cyclical in some, but pro-cyclical in other emerging economies, • inflation in emerging economies is volatile and not clearly pro-cyclical.
Table 1 shows some tentative analysis of the key macroeconomic variables in the Polish business cycle. All variables (except for inflation and net exports) were logtransformed, de-seasonalised by means of the X. 12 method and later de-trended using the HP filter. Net exports were expressed as a ratio to GDP and deseasonalised by means of the multiplicative X. 11 method before being de-trended. Volatility of cyclical fluctuations, and hence the magnitude of the business cycle, is measured by the standard deviation. Polish GDP volatility is about 0. 015, compared to 0. 08 for the euro area. The persistence of cyclical fluctuations is measured by the auto-correlation coefficient: the closer to 1, the more persistent the shock (and the longer it takes to absorb it) and therefore the less fluctuation in the business cycle. Testing for the Polish business cycle persistence yields a coefficient of 0. 55 compared to 0. 85 for the euro area. Table 1. Summary statistics of business cycle fluctuations in Poland Correlations with the cyclical component of GDP Poland Standard Autodeviation correlation -4 -3 -2 -1 0 1 2 3 4 GDP 0. 015 0. 550 1
Industrial production 3. 317 0. 730 -0. 018 0. 154 0. 314 0. 515 0. 771 0. 599 0. 444 0. 229 0. 113 Private consumption 0. 013 0. 269 0. 015 0. 079 0. 283 0. 433 0. 434 0. 385 0. 463 0. 380 0. 037 Government consumption 0. 020 0. 001 -0. 200 0. 007 0. 230 0. 022 -0. 211 0. 175 0. 282 0. 199 0. 051 GFCF 0. 070 0. 717 0. 357 0. 399 0. 403 0. 484 0. 824 0. 585 0. 371 0. 313 0. 344 Inventories 1. 189 -0. 072 -0. 139 0. 044 0. 021 0. 074 0. 199 0. 187 0. 150 0. 177 0. 230 Net exports 0. 012 0. 574 -0. 061 -0. 174 -0. 386 -0. 373 -0. 326 -0. 513 -0. 497 -0. 389 -0. 452 Exports 0. 58 -0. 005 0. 098 0. 199 0. 177 0. 225 0. 594 -0. 018 0. 030 -0. 016 -0. 193 Imports 0. 063 0. 378 0. 097 0. 252 0. 350 0. 408 0. 672 0. 377 0. 384 0. 258 0. 175 Employment 0. 017 0. 857 0. 291 0. 372 0. 411 0. 452 0. 545 0. 409 0. 333 0. 355 0. 344 Labour productivity 0. 354 0. 512 0. 120 0. 090 0. 056 0. 070 0. 133 0. 257 0. 109 0. 026 -0. 142 Money supply (M1) 0. 041 0. 776 -0. 078 -0. 060 0. 092 0. 260 0. 475 0. 472 0. 535 0. 505 0. 432 Inflation 1. 819 0. 779 0. 092 0. 321 0. 511 0. 596 0. 473 0. 290 0. 082 -0. 082 -0. 146 CPI 1. 737 0. 777 0. 086 0. 255 0. 411 0. 422 0. 00 0. 082 -0. 075 -0. 167 -0. 184 Real interest rate 0. 655 0. 942 0. 226 0. 279 0. 302 0. 263 0. 135 0. 132 0. 120 0. 092 0. 088 Source: Own estimates. Data source: Eurostat. Sample: 1995Q3 – 2007Q1 Summary statistics Lags Leads Business cycles in Poland – what is different and what is alike The analysis of the Polish cycle yields a number stylised facts, which are to some degree typical for emerging economies. Table 2 summarises a number of stylised facts on business cycles in mature economies and in Poland (bearing in mind the relatively short time p for the latter).
It shows that some of the usual characteristics of business cycles in mature economies (or even in emerging economies) are not seen in Poland. Where this is the case, some interpretation is offered. Industrial production is usually pro-cyclical and coincident in both mature and emerging economies, but in Poland it has a slightly leading property, which indicates the importance of industrial production as a driver of the business cycle. In the aggregate demand components, private consumption seems to be procyclical in Poland.
However, it is not coincident as in mature economies, and has a lead-lag profile that is not typical: it is almost flat over four quarters with some lead. Hence, although private consumption is the largest component of GDP, the dynamic relation over time between the two variables is erratic, possibly indicating consumption smoothing, which is characteristic of low-income economies. However, the ratio of the standard deviation of private consumption to the standard deviation of GDP (by which consumption smoothing is usually judged) is estimated at 0. 8, i. e. higher than the upper bound of the normal range reported in the literature. 3 This suggests that consumption smoothing is not present in Poland, which might indicate lower risk aversion and/or underdevelopment of financial markets. The Polish business cycle displays some characteristic properties Business fluctuations in Poland are highly volatile and persistent ECFIN Country Focus Volume IV, Issue 9 Page 3 Table 2. The ‘stylised facts’ about business cycles in mature economies and Poland Variable Mature economies* Poland** category Variable
Direction Timing Direction Timing Supply side Industrial production pro-cyclical coincident pro-cyclical coincident/leading Private consumption pro-cyclical coincident pro-cyclical erratic Government consumption pro-cyclical erratic counter-cyclical /erratic erratic GFCF pro-cyclical coincident pro-cyclical coincident Inventories pro-cyclical leading pro-cyclical leading Net exports counter-cyclical coincident/ lagging counter-cyclical erratic Exports pro-cyclical coincident pro-cyclical coincident Demand components Imports pro-cyclical coincident pro-cyclical coincident
Labour Employment pro-cyclical coincident pro-cyclical coincident/lagging market Labour productivity pro-cyclical leading pro-cyclical leading Money supply pro-cyclical leading pro-cyclical leading Monetary Inflation pro-cyclical lagging pro-cyclical lagging variables Real Interest rates a-cyclical erratic pro-cyclical lagging * Features commonly found in the literature. Main source: Snowdon ; Vane (2005) ** See Table 1 Source: Own calculations (see Table 1) and Snowdon ; Vane (2005) Government consumption seems to be neither systematically counter- nor procyclical.
High volatility and a very low persistence (i. e. frequent fluctuations of sizeable magnitude) point to an irregular pattern of government consumption, suggesting an important role of discretionary fiscal policies, which is a distinctive feature of emerging economies (Carmignani, 2005) and possibly related to the existence of a political cycle in public finances. Gross fixed capital formation appears to be coincident, highly pro-cyclical and persistent (auto-correlation of 0. 2) and inventories behave according to the ‘stylised facts’: they are pro-cyclical and leading (though less so than in mature economies). Finally, both exports and imports can be seen as pro-cyclical, but imports slightly more so than exports, which is in line with the features of mature, but not emerging economies where exports are a-cyclical on average. Moreover, imports seem to be quite persistent (following persistent GFCF), unlike exports (which depend on external demand); both variables are highly volatile (standard deviations are more than the double of the euro-area).
Net exports are moderately counter-cyclical with an erratic pattern over time (due to persistent imports), whereas in mature economies net exports are also counter-cyclical, but usually coincident or lagging. With respect to the labour market, employment shows up as pro-cyclical and coincident, with some evidence of lagging, which places Poland somewhere in the middle between mature and emerging economies in this respect. Labour productivity appears to be pro-cyclical and leading, in line with the ‘stylised facts’, though the leading property is less pronounced than in mature economies.
Pro-cyclicality of employment with its smooth and slightly lagged correlation profile suggests labour hoarding (Burnside et al. , 1993). Among the monetary variables, the money supply apears to be pro-cyclical and leading. Inflation seems to be pro-cyclical and to follow GDP as in mature economies, which is not the case in most emerging economies (where inflation is acyclical – see Carmignani, 2005). This is probably due to the fact that at the beginning of the transformation process all emerging economies had very high inflation rates that systematically decreased.
Thus, it is hard to discern a clear cyclical pattern. Poland managed to achieve relatively low inflation sooner than other emerging countries, allowing this pro-cyclicality to be revealed earlier. Contrary to the ‘stylised facts’ for mature (but also emerging) economies, where real interest rates are a-cyclical with no clear pattern with respect to timing, real interest rates in Poland show up as being pro-cyclical and lagging, implying a countercyclical monetary policy (coefficient of correlation with GDP is 0. 4 and there is a clear lagging pattern). Smooth and lagged correlation profile of pro-cyclical employment suggests labour hoarding The irregular pattern of government consumption suggests that discretionary fiscal policies play an important role ECFIN Country Focus Volume IV, Issue 9 Page 4 Troughs, peaks and the drivers of growth The first economic cycle since the beginning of economic transition lasted about 10 years, with the expansion and slowdown phases each pning about 5 years; the cycle ended in the second quarter of 2001.
The current upswing has already lasted 6 years, which suggests that the peak is imminent if the length of the current cycle is similar to the previous one. Decelerating leading variables (industrial production, net exports and labour productivity) may also be signs of a turning point. Chart 1. Developments of basic economic variables in Poland in 1991-2008 -10 -8 -6 -4 -2 0 2 4 6 8 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 % 0 5 10 15 20 25 30 35 % CA deficit (% of GDP) Unemployment rate (rhs) CPI inflation (rhs) GDP growth Fiscal deficit (lhs)
Source: Commission services In both the early 1990s and in 2001, when the economy was at its trough, there was a large unutilised labour supply and substantial reserves in enterprises’ capacity utilisation. The fiscal deficit was also substantial. In the first cycle, ample supply of resources was a consequence of transition to a market economy which caused a recession in 1990-1991. In 2001-2002 it followed from the world economic slowdown and a restrictive monetary policy, which forced Polish enterprises to reorganise to raise their competitiveness and efficiency. Poland got out of the trough twice hanks to exogenous impulses to investment: the restructuring of the London Club debt in 1994 (which brought the first major wave of FDI) and entry into the EU in 2004 (which led to an inflow of EU funds). The external circumstances were also favourable as the world economy expanded in the nineties until 1998 and has been on a stable growth path since 2003. Chart 2. Contributions to GDP growth in Poland in 1991-2008 -8 -6 -4 -2 0 2 4 6 8 10 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 % total consumption GFCF Inventories Net exports GDP growth
Source: Commission services The main factor that drove real GDP growth in 1995-1997 to about 7% was dynamic domestic demand growth of 8? % on average, supported by robust consumption (about 5 percentage points contribution to GDP growth on average). Despite the fact that the economy was growing above potential and domestic demand was growing significantly faster than GDP, a strong zloty appreciation (with a temporary blip in 2000) ensured that the disinflation process was not disturbed. However, this There are indications that the Polish business cycle has approached a peak Large acroeconomic imbalance during the troughs ECFIN Country Focus Volume IV, Issue 9 Page 5 situation led to a fast-growing imbalance on the current account: the 2% of GDP surplus in 1994 fell to a deficit of 6% of GDP in 1999 (to which the Russian crisis also contributed). Growth outlook based on solid foundations for now It is estimated that the Polish economy is currently expanding at a pace close to its potential (approximated at 5. 9% in 2007), on the back of rising domestic demand, which is likely to be followed by increasing imports and a deteriorating current account balance.
The growth of gross fixed capital formation accelerated to 17% in 2006 and to nearly 30% y-o-y in the first quarter of 2007. Private consumption increased at 5. 2% in 2006 and stepped up to 6. 9% y-o-y in the first quarter of 2007, which is significantly higher than the 2. 7% average growth in 2000-2003. So far, this expansion of consumption has not led to a rapid increase in the current account deficit (which rose from 1. 7% in 2005 to 2. 3% of GDP in 2006) or a rise in inflation (which has come in below the central bank’s medium-term inflation target of 2. % for eight quarters in a row), mainly thanks to moderate real wage growth. However, with a tightening labour market, emerging skill mismatches and workforce emigration the pressure on wages is expected to become more significant, contributing to a further increase of consumption. In consequence, it might lead to an escalation of the external imbalance and/or increased inflation. Nevertheless, GDP growth in the current phase seems to be based on more solid foundations than in the late 1990s: • Firstly, the share of exports in GDP has nearly tripled in 1992-2006 to about 40%; the number of exporters has also increased considerably.
The structure of exports has improved, with a bigger share of processed goods and a higher value added. Foreign direct investment has helped increase the production capacity of the Polish economy, which enables the domestic market to better meet increased private demand, and makes the balance of payments less prone to fluctuations in domestic demand. In addition, increased investment-driven imports are largely balanced by increased exports on account of a good situation in the external environment. • Secondly, the floating exchange rate is likely to act as a buffer against imported inflation.
However, even without a strong zloty appreciation, inflation in Poland during the coming years is expected to stay relatively low: below or around the central bank’s medium-term inflation target of 2. 5%. As the Polish economy is now more open than 10 years ago, inflation is more influenced by global factors. Increased exposure of Polish enterprises to international competition limits their ability to freely increase prices and wages. They are forced to increase labour productivity faster than wages to maintain their market position. Thus, even with growing wage demands, enterprises are more willing to decrease mark-ups than to raise prices. Thirdly, increased household incomes acquired as a result of higher wages and an improved labour market situation may not translate into consumption to the same extent as in the previous economic cycle. There is evidence that households are now more saturated with basic durable and consumption goods which they lacked before4 and are more eager to spend additional income on holidays abroad owing to a more mature service sector. Financial markets are more developed than 10 years ago, giving an opportunity for financial investments. Conclusions
The business cycle in Poland exhibits similar properties to cycles in mature economies, but there are some notable differences for government consumption, net exports and real interest rates (although for the last variable the picture may be blurred by its very high level at the beginning of the transformation process). However, because the data series are short, the results should be interpreted with caution. The irregular behaviour of government consumption in Poland with respect to influence on the business cycle could be related to a discretionary fiscal policy implemented within a political business cycle.
GDP growth is based on more solid foundations in the current cycle ECFIN Country Focus Volume IV, Issue 9 Page 6 The analysis of the previous upswing in Poland, the identification of variables with leading properties with respect to GDP, and the latest developments all seem to suggest that the Polish economy might have reached the peak of the current cycle in the first quarter of 2007. Nevertheless, thanks to the ongoing process of restructuring of the economy, the slowdown phase is not likely to be as pronounced as in the previous cycle and should not lead to major imbalances.