POSSIBLE IMPACT OF THE 2008-2009 ECONOMIC CRISIS ON CZECH POTENTIAL OUTPUT THROUGH THE LABOUR MARKET

The paper deals with the possible impact of the 2008-2009 fi nancial and economic crisis on the potential output in the Czech Republic. Given the general diffi culty in sketching out all the consequences of the crisis itself, the article is primarily focused on the labour market, which is regarded as an important driving force in terms of potential output fall or its growth slowdown. First, principal reasons why the potential output could have been hit by the 2008-2009 crisis are discussed in detail. The paper then analyzes a number of transmission mechanisms through which potential output could be generally impacted by severe recessions. Further, the output gap of the Czech Republic is estimated and a potential output development is drawn up. Finally, an approximate impact of the crisis on the potential product through the labour market is demonstrated. The results show that the potential output in the Czech Republic has been partially and temporarily affected in volume. This is specifi cally due to an adverse long-term unemployment and, to a much lesser extent, employment behaviour, in both cases seemingly in relation to the recent crisis. Likewise, Okun’s law has been found to be relatively weak in the case of the Czech economy, thus further supporting the “hysteresis explanation”. However, such a drop in potential output will likely not persist in the long run, provided that appropriate economic policies – in particular aiming at re-integration of long-term unemployed/ discouraged workers and an increase in the employment and participation rates − are implemented.

fi city of market structures into one model, thus making use of the microeconomic foundations principle.
Technically, however, the majority of economists tend to use the typical Cobb-Douglas production function. Moreover, this function has also served as a preferred tool of international institutions such as the OECD or the ECB for many years in row. It is perhaps the following form which is exhibited most often: where Y t stands for potential output; 1 t L  is potential/full employment in power of its proportion in the respective economy, while is full employment given as workforce less natural rate of unemployment (or NAIRU). Finally, t K  stands for stock of capital over its proportion in the respective economy. 3 However, it should perhaps be noted that the above outlined mainstream approach is completely opposite to what some other economic streams suggest. For instance, in some of their fi ndings and interpretations, Post Keynesians state that potential output is always lower without nominal rigidities than with them. 4 Hence, a relatively high level of regulation is to be maintained for an economy to work at its real potential output. Moreover, they argue potential output with nominal rigidities is also much more stable, which essentially means a lower unemployment level (and a higher employment level respectively) is kept more steadily than the lower unemployment level constantly requiring higher levels of government interference. All this reasoning relates primarily to the effective demand problem. According to Post Keynesians, the output is a function of aggregate demand even in the long run. 5 Similarly, unlike mainstream economists and/or Post Keynesians, a number of heterodox economic scientists question the very usefulness of potential output calculations. Among them we can fi nd Austrian School adherents, who generally believe that output gaps are overestimated as per rule, since the whole business cycle amplitude is caused by central bank (activist) monetary policy. This happens when potential output is measured as an average of past economic performance on a trend-fi tting basis. In other words, when determining the position of an economy within its business cycle, central banks take into account past conjunctures that were artifi cially caused by monetary stimuli. Subsequently, when this "artifi cial" boom is over and malinvestment capacities are being destroyed and/or abandoned, it is not (effective) demand but rather structural factors that bring the recession about. Hence, the economy de facto still works on its potential − even during the recession phase (Hayek, 1967).
3 A similar expression, where "N" denotes the pool of workers and equilibrium employment is given as * . U u N   , can be found, for instance, in Vošvrda (1996) in the form of   1 * .

Y K N U
Using the theoretical framework for potential output described above, let us turn our attention to the recent economic situation. According to the latest data, in 2009 the Czech economy -as many others in the European Union -emerged from the recession that it was hit by in the course of 2008 and 2009. 6 In the Czech Republic, the (technical) recession started in the fourth quarter of 2008 and ended in the third quarter of 2009. Put differently, the country was technically experiencing recession for three quarters of one year. However, doubts still exist concerning two principal aspects of the past crisis: fi rst, whether it could occur again in similar magnitude in the foreseeable future. Such doubt is notably attributed to the fact that governments in developed countries are currently being forced to limit their expenditures due to high public debts and, more generally, deteriorated public fi nances in terms of their structure and revenue-spending balance. Secondly, one could question whether or not potential output has been affected as a consequence, and if so, to what extent. The second question should be at least partially answered in this paper.
Whilst the crisis might have likely led to some material long-lasting shrinkage in potential output levels, there is also a risk that it could forestall potential output growth from returning to the pre-crisis rates in the medium to longer run. As a consequence, the principal aim of the paper is to examine whether the 2008-2009 fi nancial and economic crisis has had any substantial incidence on potential product, and therefore showed signs of acyclicity. Since the crisis is still somewhat present in the economy -mainly in the form of anaemic output growth and higher rates of unemployment in a number of developed economies -we focus on the qualitative (i.e., explicatory reasons) and quantitative (in terms of potential GDP losses in the Czech Republic) aspects of possible potential product (growth) decline/slowdown.
Essentially, to answer such questions we need to estimate potential GDP itself. As it has been already mentioned, there are several techniques for calculating it, most of them broadly recognized by the contemporary mainstream economic science. Most of these techniques focus on calculating potential output upon the Cobb-Douglas production function (Hurník, 2005). However, as it is widely perceived, apart from many benefi ts this approach brings about, it also has several inconvenient facets. Among those should be highlighted the notorious "trend problem", typical for population growth and capital accumulation, and the not-that-accurate capital stock estimates, which are usually made retrospectively. Yet, as some economists are ether skeptical to such calculation because of the problems with total factor productivity (TFP) aggregation, they prefer to calculate potential product based on average GDP growth. Such an approach is presented in this article as well.
The rest of the paper is organized as follows: the fi rst section draws up on several transmission mechanisms through which recessions can likely affect potential output. This is analyzed mostly by generalizing previously available results for developed economies. In the second part, we calculate the potential output of the Czech Republic using output gap and employment gap estimated by Hodrick-Prescott fi ltering; and in the fi nal part, we present results focusing on the Czech labour market, including helpful fi ndings for the dynamized gap version of Okun's law.

Economic crisis and potential output
The intensity of the fi nancial and economic crisis of 2008 and 2009 was to a large extent unexpectedly strong and unprecedented at least for several past decades. It is sometimes even admitted the crisis/recession was the strongest crisis since the infamous Great Depression of the 1930s. Moreover, a number of authors suggest fi nancial crises are more likely to become crises affecting even potential output. According to diverse sources, the incidence on potential output of several fi nancial crises before the 2008-2009 one is estimated to around 1.5% − 2.4% on average 7 (Lemoine and Pavot, 2009).
We might fi nd a number of good reasons why potential output is suspected to have been hit by the recent economic and fi nancial crisis. Firstly, the crisis might have diminished potential output due to the negative effect on investment, as the disturbed capital accumulation might have resulted in increasing obsolescence of the capital stock. In this sense, we shall argue that the number of companies going into liquidation has mainly impacted the short-term capital stock, and therefore short-term potential output behaviour (EC, 2009). Also, larger write-offs and faster depreciation rates, which are often caused by the need for stock reclassifi cation, contribute to the matter 8 (Lebrun, 2009).
Secondly, labour force could have been affected since the human capital accumulation and restoration might have been suspended. Finally, the total loss of labour force might have been even higher with a higher number of people discouraged and therefore quitting the labour market as such. This could have been achieved by requesting premature retirement, or, for instance, moving to the shadow economy labour force, and thus generally lowering participation rates and overall employment.
Further, behaviour of the private sector seems to have been exaggerated or excessive in terms of fi rms' reactions and expectations; in recessions, company managements tend to be overly pessimistic of the future, and they lay off more labour force than would be "reasonable" in the context of existing consequences. We call such behaviour "unreasonable" as it is typically revised when the economy gets out of the recession. On the other hand, businesses usually keep the "core" labour force consisting of workers that they consider essential for their base functioning. Such workers are, however, usually the most costly ones, for they are not only the best paid workers in the company, but they are usually granted high bonuses and extra-salary compensations. However, their short-term marginal as well as average propensity to consume is lower compared 7 Among the developed countries experiencing this kind of problems we could single out Spain in 1978-1979, Norway in 1987-1988, Finland in 1989-1993, Sweden in 1991-1993, and Japan in 1992-1993 Essentially, standard macroeconomic theory recognizes three ways how capital intensity can evolve: 1) capital deepening; 2) capital widening; and 3) capital shallowing. However, we are still not sure whether the crisis brings about capital shallowing or capital widening effects. This is due to lack of trustworthy empirical evidence.
to those who have been laid off, which means their contribution to the GDP growth is relatively lower (per unit of income). 9 Moreover, once the economic pressures arise, they might fi nd "fertile land" in politics. The general public typically force politicians to adopt measures that contribute to the ease of potential performance. Among these measures there are, for instance, a possibility of early retirement, barriers to migration fl ows, stricter work permits for foreigners, higher unemployment benefi ts, foreign trade restrictions, etc. All these measures normally lower the level or growth of potential output, because they usually lead to a decrease in the participation rates (which is especially true of the most affected cohorts) as well as the overall employment rate.
Most importantly, hysteresis frequently takes place in the form of the insider-outsider effect and/or the duration effect (human capital losses). The insider-outsider effect generally refers to a situation where trade unions or any other interest groups act in favour of labour market incumbents. This is most frequently done in the form of wage negotiations which fi nally lead to lower employment of outsiders when the recession is over. 10 The duration effect is different: it refers to a situation where long-term unemployed become hardly "employable" at all, as their losses on human and social capital are perceived by possible employers as insurmountable. Moreover, in the course of time these unemployed also frequently become discouraged, which means that they completely give up searching for any job opportunities. The hysteresis effects have been very precisely analyzed by Olivier Blanchard and Lawrence Summers in their 1986 article called Hysteresis and European Unemployment. Since then, many other estimates have confi rmed that European countries generally face higher dangers of their (long-term) unemployment rates turning into hysteresis than the OECD average (De Lucia, 2010; Reinhart and Rogoff, 2009).
Concerning the two aforementioned hysteresis effects, we may also mention that they should not be confused with so-called policy inertia on the labour market. Although the two are necessarily interrelated, policy inertia takes form of a labour market regulatory framework, such as minimum wage legislation, income taxes, social and healthcare contributions, unemployment benefi ts, working conditions or probation and notice periods. Unlike in the case of hysteresis, all these measures are put into effect intentionally, as a part of country's respective economic policy.
Finally, a very profound impact on the level of potential output is to be felt through the investment channel. The unprecedented drop in investment activity brought about by this crisis is likely to be present in the economy for quite a long time. This is connected with the pessimistic expectations of entrepreneurs concerning future prospects of the economy. Such expectations usually lead to a change in attitudes towards risk taking, making it more constrained and, in extremis, less likely to appear. Moreover, credit constraints and behaviour of the banking sector are expected to play their part 9 Such a statement is valid namely in recessions and/or for more considerable output slowdowns, where the mentioned propensities are typically low in general. 10 However, many additional illustrations of the insider-outsider effect can be singled out even without trade unions or interest groups playing part in them.
as well. Hence, restructuring of production will end up being much more costly than would be desirable. Likewise, as fi rms do not have suffi cient access to credit, they tend to get entrenched in less productive capacities, not being able to employ their resources in more effi cient production possibilities. Even here, hysteresis effects play their role in the form of capital and credit granting hysteresis 11 (similar and more elaborated arguments are listed in EC, 2009). Apparently, the reaction of the economic policy targeting potential output is extremely important as it should aim at stabilizing or even increasing labour supply. However, the Czech economic policy is not grossly equipped for such crises not only due to its great openness, but also due to the fact it has not experienced anything similar during its post-1989 history.
Generally − no matter whether with or without proper economic policy reaction − we are now facing four plausible scenarios of potential output development: fi rstly, no loss in potential output at all will be infl icted. In other words, we could analyze the 2009 recession as not having any impact on potential output in terms of material loss/ growth deceleration. Such a scenario, however, is not very probable (reasons suggested, e.g., in Blanchard, 2009, or in Reinhart andRogoff, 2009) given the presented context. The second scenario works with a temporary loss in potential output, forecasting that it will ultimately -in a matter of a few years -return to its pre-crisis path. We could also call it a "full-recovery" scenario. The third one suggests there will be a permanent loss in potential output in the wake of the crisis, but its growth will remain unaffected. Finally, the worst possible scenario works with a "widening gap", caused by a slower pace of potential output growth in the long run. These four scenarios -together with "their" simplifi ed corresponding growths -are exposed in the following charts.

Behaviour of components: unemployment and real GDP
We have collected available quarterly data for the Czech Republic over the period of 2006-2013 (including a forecast for 2011-2013) and adjusted the in a way that they fi t into the following HP fi lter. At the fi rst glance, it might appear that the scenario for the Czech Republic would perfectly match the scenario called "one-off decrease". In any case, the analysis would have been made more inaccurate if we had kept an argument based solely on fi ltering. Source: own elaboration based on data from CZSO and MoF.
In terms of assessing an impact on potential output, we have opted for the labour market as a driving force, holding other components constant. This session presents the results for the Czech Republic that were obtained by time series analysis.
Having used the Hodrick-Prescott fi ltering (parameter λ = 1600, which is conventionally used for quarterly observations) and given the data for the year from Czech Ministry of Finance (Macroeconomic Forecast, 2011), we get the following results: while Fig. 6 shows smoothed unemployment levels from 1993 to 2011, Fig. 7 illustrates how much it was actually affected in terms of percentage point changes.  Smoothed changes in ILO unemployment rates 1993ILO unemployment rates -2010in p.p.) Source: own elaboration based on data from CZSO, 2011.
Concerning the Czech labour market data, we could probably stress a couple of relatively important observations: given the gravity of the fi nancial and economic crisis, long-term unemployment increased by approximately 2 percentage points over the period 2009-2010. According to general literature (e.g., Blanchard and Summers, 1986;Elmeskov and Pichelman, 1993), hikes in long-term unemployment always lead to a one-off decrease in potential output because of its adverse effects on skills and work and/or social habits of the people concerned. Many authors suggest, and it has repeatedly been demonstrated by empirical studies (e.g., Llaudes, 2005), that it is not solely a problem of productivity that prevents re-employment, but also a problem of labour demand. Employers/fi rms are often rather reluctant in employing labour that has been unemployed for a relatively long period. Most importantly, such reluctance is frequently caused by higher real costs of labour force caused by its "wasting". Such costs include re-taken social and work integration processes, training expenses, uncertainty and unpredictability concerning employees' behaviour, loss of motivation to work caused by long-lasting unemployment, and, once employed, limited labour availability for an employer due to the shorter period to remain in available workforce. In addition, long-term unemployed are often very much stigmatized by employers without any real substance. As in other cases, even here we can see a certain mixture of justifi able and reasonable behaviour with one which is rather emotional and/or subjective.

Long-term unemployment
According to the 2011 survey made by the Czech Ministry of Finance, the economic crisis appears to have led to moderately lower utilization of workforce (Macroeconomic Forecast, 2011). 14 Such utilization is by and large a coincident indicator and typically fi nds its roots in labour hoarding effects, which lead businesses to keep part of their workforce, so that they do not have to hire it again once the recession is over. Also, since laying off redundant workforce, re-hiring and (re-)training are costly exercises, businesses may be inclined to wait some time to see how the situation evolves. This may be seen as one of the reasons why the unemployment rate is conventionally considered a lagged variable (Millard, Scott and Sensier, 1997).
In the course of the fi nancial and economic crisis, both unemployment and long-term unemployment have been experiencing very different scenarios as is demonstrated in the table below: Apparently, even though the long-term unemployment increased considerably in the wake of the crisis, its share in the total unemployment decreased in 2009. This fact is essentially driven by purely "technical" reasons as the laid off workers cannot simply become long-term unemployed right from the beginning of their unemployment "regime". Referring to the above table, one can clearly see the long-term unemployment has shown a rather increasing trend in recent years. This is very likely a consequence of the economic recession as the most considerable hike was in 2010 (increase by approximately 44% compared to 2009).
While the spells of long-term unemployment are namely apparent in human capital destruction and technically in a hike in the NAIRU, there is yet another problem impeding labour market to return towards equilibrium. Essentially, in the long run, "incumbents" on the labour market (i.e., people currently employed) exert substantial advantages compared to "entrants", which in other words mitigates the dampening effect on the wage claims the entrants normally have. This also prevents entrants from being hired by the companies. Such a process has been put forth by a number of modern economists focusing on labour market conditions, among whom Blanchard and Summers (1986) are perhaps the most cited ones.
Judging from the above table, we can see there was a constant decrease in long-term unemployment beginning in 2004, while over the period 2004-2008 the amount of long-term unemployed more than halved. Moreover, without some kind of path dependence, which could have been present in the Czech economy to a certain extent for approximately 13 years since the vast banking crisis in 1998, this decrease could have been even more pronounced. Also, the driving forces most probably lay in the cuts in unemployment benefi ts (in both amount and length), deemed to be directly affecting the "willingness" to remain unemployed. Besides this, the general rate of unemployment was steadily decreasing as well, which partially led to a drop in the share of the long-term one.
Long-term unemployment is commonly defi ned as an unemployment lasting more than 12 consecutive months (Eurostat, Czech Statistical Offi ce, etc.); however, in some countries like the U. S., only 27 weeks (i.e., half a year) suffi ce for an unemployment period to be classifi ed as long-term (US Bureau of Labour Statistics). In the Czech Republic, this type of unemployment increased substantially soon after the beginning of the crisis, and continued to rise even after the recession subsided. In 2010, it reached approximately 31.8% of all the unemployed people. Such a high number is also typical of most European countries as their hysteresis effects are considered more signifi cant than those of the US and Japan (De Lucia, 2010). Most often, the less fl exible labour market conditions in the European countries are typically taken for the main culprit. On the other hand, this theory is now put in question since the general persistence of unemployment is currently higher in the US than elsewhere. Additional research on the impact of fl exible labour market conditions should thus be performed after the crisis is completely over.
For the purpose of demonstration, we can construct a simple model in which we investigate how much of the (short-term) unemployment was transformed into long-term. For this reason, we have estimated a dynamic equation in the following form 15 : where LTU stands for the long-term unemployment and U for total (general) unemployment. In the Czech Republic, it seems there is empirical evidence that the proposed relationship holds true. Taking 42 observations (n=42) from 3Q 2000 to 4Q 2010 fi ltered through the HP fi lter and using standard λ = 1600, the simple OLS model for one-quarter lagged LTU and U reveals the following: Hence, according to the results, there are still signifi cant hysteresis effects on the Czech labour market, which essentially means a certain number of the unemployed eventually end up being long-term unemployed. Although there have been many attempts to counter this pattern, the economic crisis has made the problem more accentuated once again. It is very likely that through this transmission mechanism the 1 6 possible impact on potential product might have been the most important. 17 Figure 8 then illustrates the relatively strong relationship between unemployment and long-term unemployment. In other words, the Czech labour market tends to be experiencing more intensive carryover from normal unemployment into long-term than is typical for other developed countries. According to what has been presented above, even though the Czech Government has been trying hard to lower the rate by cutting unemployment benefi ts and generally tightening up the eligibility (which has admittedly been successful to an extent), the economic crisis has put a halt on such efforts. Certainly, more research will be needed to fi nd out whether signifi cant levels of long-term unemployment are still present in the economy several years ahead of the end of the crisis. Changes in unemployment rate vs. changes in long-term unemployment rate (y-o-y; in %) Source: own elaboration based on data from CZSO, 2011.

Yet another variable possibly infl uenced by 2008 and 2009 crisis is employment.
Overall employment and participation rates 18 are in general handy economic variables as they can play a double role of a cyclical indicator and at the same time exert an impact on potential output. Most economists assume, which could also be accepted as conventional wisdom, that potential output is likely to be higher in countries where overall employment and general economic participation are higher and vice versa (e.g. Lemoine and Pavot, 2009).
In the case of the Czech Republic, the labour force (given as the number of people aged 15 to 64 who are either employed or unemployed) was nearly constant during the crisis and right before it, which means the conclusions can be drawn with respect to the employment rate only (Macroeconomic Forecast, 2011). On the other hand, it could well be expected that the 15-64 population share in the whole population will be declining over time as demographic pressures start to be felt more deeply. 19 18 Participation rate conventionally depicts available workforce -employed plus unemployed − over the population aged 15+. Sometimes it can also be defi ned as the available workforce over the population aged 15 to 64. 19 According to the statistics, we should reckon on a slow decline in population aged 15-64 already in 2012 − by 0.8 per cent year-on-year (Macroeconomic Forecast, 2011).

Figure 10
Changes in employment rate, 1994-2010 (y-o-y; in p.p.) Source: own elaboration based on data from CZSO, 2011. In order to compare the results for long-term unemployment and employment, let us examine the impact of the economic slackdown on the employment fi gures. We can estimate such an impact through the following equation: Taking 42 observations (n=42) from 3Q 2000 to 4Q 2010 fi ltered through the HP fi lter and using standard λ = 1600, we can see the model is, based primarily on the Akaike information criterion, signifi cant for the lag of two quarters. However, the adjusted R-squared is not so convincing: it equals only 0.6, therefore showing a relatively mesoscale relationship. Once we focus purely on the period of the crisis itself, i.e., from 2Q 2007 until 4Q 2010 (n=15) it might be seen the model proves signifi cant for a lag of one quarter. Again, this is based primarily on the Akaike information criterion. Essentially, this suggests that the proliferation of the slowdown on employment follows a typical relationship between GDP growth and unemployment as at the same time the ILO rate of unemployment reached 7.3% at the beginning of 2010, and happened to be the highest since 2004. In other words, we are once again testing the following equation: 20 (4) In this case, the R-squared proves to be very high, explaining more than 90% of the variability. Such a relationship might be caused chiefl y by the short time series used; however, it points in the direction of a somewhat more signifi cant tie between the two variables with the economic crisis coming in. The proposed models are not entirely able to show any considerable incidence of GDP slowdown/drop on employment, and therefore empirically proved the expected link between employment and potential output.
Besides, looking more closely at the participation rates and the unemployment rate behaviour, while abstracting from the long-term one, it seems that employment has predominantly played the role of a cyclical indicator instead of one affecting potential output. We can thus assume the number of discouraged workers, quitting the labour market because of the economic crisis, has been rather low. Therefore, analyzing and comparing the two results, we cannot draw any strong conclusions in either way. 21

Okun's law and hysteresis
Finally, we are going to examine the contemporary usefulness of Okun's law in the case of the Czech Republic. The following analysis enables us to examine the possible link between unemployment and output gap. In other words, we will try to determine whether or not the changes in the cyclical component of the unemployment rate can be explained by the real output development. Let us defi ne the dynamized gap version 22 of Okun's law as follows: with the cyclical part of real output is considerably weaker (-0.19). On the other hand, the inertia for the cyclical component of the real output happens to be substantially higher (0.42).
These results underpin the evidence that the link between cyclical movements in unemployment and output are, in the case of the Czech economy, weaker than those suggested by A. Okun himself. Equally, our results are broadly in accordance with some cross-country studies, such as Železník (2010). These fi ndings prove that the impact of recessions may not only be consequently visible in cyclical components of the Czech output, but could be partially transformed into a persistent part of it in the form of hysteresis on the labour market. For this reason, the general validity of Okun's law is put in question in the context of the Czech Republic, while the proliferation into the structural parts of the economy is further supported. (Figure 11 sketches in the mentioned rationale by the impulse-response reaction, simulating ten periods ahead of the shock.)

Figure 11
Impulse-response reaction of cyclical unemployment on cyclical output with 95% confi dence band (in grey)

Conclusion
The main body of the article has attempted to examine whether the potential output in the Czech Republic might have been hit by the 2008-2009 economic crisis, primarily judging from the labour market evolution. This could essentially happen as a result of two aspects: fi rst and foremost, because of the specifi city and particular proliferation of the crisis, but also, secondly, given its almost unprecedented severity. Overall, we may conclude as follows: on the one hand, it seems that the potential output has been partially affected due to the persistent hysteresis effects in the Czech economy; on the other hand, we have to admit that the most important measures on the labour market do not signal any extreme disturbances. This is especially true comparing the development with the pre-crisis period. Nonetheless, a more signifi cant impact on our potential product might be visible in the years ahead once longer and more accurate time series are available. Based on the entire analysis, whilst referring to the sketched scenarios depicted by Figures 1-4, we would therefore favour the one-off decrease in the potential output. This is most importantly with respect to the long-term unemployment spike, a moderate fall in the employment levels, and Okun's law insignifi cance, all of them essentially signalling that labour market hysteresis effects may still be present in the economy.
Regarding a proper economic policy response, together with Lemoine andPavot (2009) and/or De Lucia (2010) we can argue that the most important one is to abstain from pro-cyclical fi scal policies, while at the same time implementing suitable and sustainable structural ones. More than ever this is a very important economic policy recommendation. To be more specifi c, any such economic policy should chiefl y target long-term unemployment, insider-outsider issues, and lack of proper human capital investment and training activities. In contrast, not all the recommendations allegedly targeting structural problems in the economy are to be considered as necessary. This holds even truer when economic context and country specifi cities are not properly taken into account.
Nonetheless, we do expect, once the fi nancial and economic crisis is completely over, more thorough and rigorous conclusions about potential output development to be made, although their relevance for economic policy recommendations will be apparently much weaker. Therefore, their major role rests in a possible "lesson learnt" for the future. Moreover, they can prove right or wrong all the attempts to estimate the impact more upfront than might actually be desirable in terms of accuracy.
Finally, a more signifi cant impact on the potential output in the Czech Republic is to be expected with the (adverse) demographic changes that are still a few years ahead. Provided that no appropriate economic policy measures are adopted, the potential output will likely be affected not only in volume, but also in terms of its growth.