Abstract
We examine the role of organized labour in the restructuring experience of two coal-mining regions in the 1990s. Under similar external circumstances, the Czech Republic’s Ostrava region underwent gradual restructuring from early on whereas Romania’s Jiu Valley region went through no restructuring until 1997 followed by massive layoffs over 2 years. We conduct a quantitative exercise that accounts for mine productivity, labour market conditions and the constraints in compensating laid-off miners. We show that the Jiu Valley’s delay in restructuring was inefficient. Gradual restructuring with compensation would have benefited both the miners and the government. The proximate reason for the delay was the Jiu Valley miners’ action against restructuring. We discuss what motivated their action and why it was effective.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
For the opposite view, that various associations can facilitate politico-economic modernization, see Putnam (1992).
- 2.
See The Economist (1999).
- 3.
The spirit of this research is shared by Galdon-Sanchez and Schmitz (2002), who study the change in labour productivity under increased competition in the worldwide iron-ore industry.
- 4.
Our model can be viewed as a version of the sectoral reallocation model (that is, Aghion and Blanchard 1994; Castanheira and Roland 2000) that studies the properties of transition paths as resources are reallocated from the state to the private sector. In comparison to the others, our model is designed for the episodes of restructuring at hand: quantitative results are sought under specific assumptions.
- 5.
We interviewed government officials, mining company managers, union representatives and researchers. Five interviews were conducted in Prague and Ostrava in 2002, and fourteen in Bucharest, Jiu Valley and Cluj in 2002 and 2003. These were complemented by previous studies, mainly sociological studies on the effects of restructuring in Jiu Valley, which are referenced throughout.
- 6.
To be precise, the Ostrava region refers to the districts of Ostrava and Karvina in northern Moravia, and the Jiu Valley region refers to the townships of Petrosani, Aninosa, Lupeni, Petrila, Uricani and Vulcan in Hunedoara County.
- 7.
These are rough estimates based on Statistical Yearbooks of the two countries and Tables 1 and 2 (see Footnote 8). According to the Czech Statistical Yearbooks for 1986–1991, coal production in Ostrava region was between 22 and 23 million tons per year from 1985 to 1989, dropping to about 20 million tons in 1990 and 1991. According to the Romanian Statistical Yearbook for 1993, hard coal production in Romania was between 11 and 12 million tons per year from 1986 to 1989, dropping to between 5 and 6 million tons from 1990 to 1992. Discrepancies between the figures for the early 1990s and those in Tables 1 and 2 may be due to a different classification of coal, different degrees of processing, inclusion/exclusion of minor regions and so on.
In Jiu Vally, arguably, there is a recovery of production in the early and the middle 1990s although the size of recovery is dwarfed by the initial decline in production from the late 1980s (see Table 2 and Fig. 2). We could speculate on various reasons such as the over-reaction in the first years, short-term fluctuations in demand and the union’s power to control production at the margin. Importantly, there is no evidence that there was a steady recovery of demand that would have continued if there had been no restructuring.
- 8.
The data for Ostrava are from the yearbooks published by the Czech Mining Institute (1993–2002). The data for Jiu Valley were directly obtained from the CNH, the only mining company in the region. We cautiously present the 1990 data for the Ostravian mines: their only source is the 1993 yearbook and the data from this yearbook are in general somewhat different from the subsequent yearbooks. For both regions the employment figures are for the workers in mines, excluding the headquarters and auxiliary units. Much of the change in the number of these excluded workers represents spin-off units (for example, social services) that are not directly related to mining. We focus on the restructuring of mines proper and abstract from this peripheral issue.
- 9.
The improvement in labour productivity over the period is largely due to downsizing. In particular, the Jiu Valley mines had a large slack in labour until 1997, given the large initial decline in output. Further, there was little investment in new technology in the Jiu Valley throughout the period. Some Ostravian mines, however, adopted new technology in the early years, which contributed to the improvement in productivity.
- 10.
The aggregate output and unemployment data are from World Development Indicators 2007, published by the World Bank.
- 11.
The labour market data were directly obtained in various parts from the statistics and labour offices at both the regional and national levels. The official unemployment figures measure only registered unemployment and seemed to underestimate unemployment (that is, the labour force as measured by the sum of employment and unemployment figures shrinking beyond immigration flows) during the downturn of the labor market, indicating fluctuations in unregistered unemployment. This problem was clear and significant particularly in Jiu Valley in the late 1990s. We estimated unemployment rates by subtracting the official employment figure from our estimate of the labour force. For the Ostrava region we fixed the size of the labour force at the average over the period since there was little population growth or immigration during the period. For the Jiu Valley region the labor force was held at the average until 1996, and afterwards adjusted year-by-year by subtracting the significant net-immigration out of the region. Finally, note the implausibly low unemployment rate in Ostrava in 1996, which is due to an obvious error in the raw data for this year. Not knowing the magnitude of the error, we left it uncorrected.
- 12.
A 1996 European Union assessment of the labour market conditions of 21 European coal mining regions ranked the Jiu Valley region as the fourth most “vulnerable” to restructuring, and the Ostrava region as the third least vulnerable (see European Union 1996).
- 13.
The average wage was about 61% of the mining wage in Hunedoara county from 1993 to 2000. The average wage in Ostrava and Karvina districts was about 77% of the mining wage, calculated as the gross wage divided by employment in OKD accounts, in the same period. (See Footnote 6 for administrative divisions of the regions.) Further, Chiribuca et al. (2000) estimate that the ex-miner’s wage was a half of the average wage in the region. However, some of this differential is probably transitory and reflects the temporarily worsened labour market conditions and the trial-and-error aspect of a job search. Ex-miners typically worked as labourers in construction, on public-works repair and on the black market (Larionescu et al. 1999).
- 14.
The hunger strike should be understood euphemistically: most participants left the protest site in the evening, presumably discontinuing hunger. Further, the participants considered the granting of governmental benefits to non-participants as encouraging “free riding,” which makes it clear that the participants perceived the strike as a legitimate means of obtaining personal benefits.
- 15.
The OKD produces about 80% of Ostrava’s coal and, as mentioned, the CNH is the only mining company in Jiu Valley. The accounts for the earlier years were not publicly available.
- 16.
To elaborate on the rationale for constancy, it is plausible that the OKD could choose output and non-labour input (but not labour input) throughout the period. Then the marginal product of non-labour input would have been equal to the input price level, which implies that the non-labour input share was equal to α in (1). The CNH, on the other hand, probably could not choose output (and labour input) freely due to its more rigid output market, leaving no prediction as to whether the non-labour input share should increase or decrease as we move from the early to the late 1990s: the 1997–1998 layoff would have increased the share of non-labour input as a substitute, but the accompanying reduction of output would have had the opposite effect. The CNH account suggests that the two effects roughly cancelled each other out.
- 17.
Note that the labour productivity Y t /L t rises during restructuring in both regions. From (2) and (3), the total factor productivity A t rises during restructuring if \( \bar A(1 - \alpha )/\gamma < \) Y 0 /L 0 , which depends on α and γ. In the case of the Ostrava region, we could set α = 0.59 (see Footnote 16) so that the condition becomes γ > 0.13, which implies that the labour cost be at least 13% of revenue if the labour market were competitive. This is a highly plausible conjecture. Similarly, if we were to set α = 0.48 for Jiu Valley although we are less sure of its validity, the condition becomes γ > 0.18, which is again highly plausible.
- 18.
For the Ostrava region, the function is drawn through the 1991 point rather than the 1990 point. The large output drop from 1990 to 1991 largely represents the initial demand shock, as mentioned in Sect. 2, rather than restructuring.
- 19.
The price was measured dividing revenue by production volume and then converting this value to USD according to the current exchange rate. If we discount the price by the Consumer Price Index instead, we obtain the same periods of instability and a somewhat downward trend over the whole period, which is expected since the real value of USD depreciated over time. We assume that the price difference between the regions reflects the quality difference. Price fluctuations reflect minor demand shocks, except when the price increased from 22 USD in 1996 to 27 USD in 1997 in Jiu Valley, decreasing to 22 USD by 2000. The reduction in production due to the layoffs may have raised the price and/or the remaining mines and coal seams may have raised the quality of coal.
- 20.
The job-finding function can be considered as a version of the matching function \( {m_t} = bu_t^\theta v_t^\vartheta \) with the free entry condition \( c = {f_t} \cdot {m_t}/{v_t} \), where \( {v_t} \) is the number of vacancies, c is the vacancy cost, and \( {f_t} \) is the value of a filled job. Setting the matching to be decreasing returns to scale (\( \theta + \vartheta < 1 \)) and fixing the value of a filled job \( {f_t} \), we obtain (4) by substitution and with the appropriate changes of variables. The same result follows from setting the value of a filled job to be log-linear in the number of matches (\( {f_t} = zm_t^\nu \) where z is a constant and \( \nu < (1 - \theta - \vartheta (1 - \theta ) - \theta (1 - \theta ))/(\vartheta (1 - \vartheta )) \).
- 21.
We could only obtain the annual data of unemployment duration for the greater Ostrava region, which is composed of five districts including the two districts of our interest, Ostrava and Karvina, from the Czech Statistical Office. Since the drop in the job-finding rate may have been more severe in the two-district region, the following estimate of ρ may be an overestimate.
- 22.
This is based on Chiribuca et al. (2000) and the annual data of unemployment duration for Hunedoara county that includes the Jiu Valley. Chiribuca et al. (2000) estimate that 27% of all miners laid off from 1997 to 1999, nearly all of whom were laid-off in 1997, were working in 1999. For the Hunedoara county, we estimated the expected unemployment duration to be about 29 months in the1999–2000 period.
- 23.
Using the annual data of unemployment duration at the national level from the Romanian Statistical Office, we estimated the expected unemployment duration to be about 12 months in the 1996–1997 period. The unemployment rate of Jiu Valley was comparable to the national average prior to 1997. Our guess would be valid if the labour market conditions aside from the unemployment rate (scale-adjusted B in (4) and (5)) were also comparable.
- 24.
The lack of precision for ρ is not crucial for the main result, as the sensitivity analysis in Sect. 3.3 shows.
- 25.
There would have been extra unemployment linked to mining in Ostrava as well, although we don’t know its size. Since the extra unemployment linked to mining would lead to a slower pace of optimal restructuring, the bias resulting from the modeled asymmetry between the regions is a safe one: it reinforces the result that the actual restructuring path in Ostrava, although slower than the optimal one, was closer to optimal than in Jiu Valley.
- 26.
Recall that there was a significant emigration out of Jiu Valley in the late 1990s, no doubt as a consequence of the massive layoffs (Footnote 11). Since \( \eta \) is calibrated to the unemployment path in Jiu Valley, its value is net of the emigration of those laid-off, including ex-miners.
- 27.
Here we have assumed that the job-finding rate of the emigrant ex-miners is the same as that of the miners who stayed (see Footnote 26). We are not sure about the employment prospects of the emigrant ex-miners, but they were probably not very different from those of the ex-miners who stayed. Many of the emigrants returned after a while, unable to find work elsewhere.
- 28.
That is, before any taxes including the social security tax paid by the employer. The gross wage is the proper measure of the market value of labour.
- 29.
Thus we are measuring the value of labour both inside and outside mines in units of coal and then multiplying it by the price of coal, which is assumed to be fixed in USD (see Sect. 3.1). Again, the fixed-price assumption avoids the nominal noises of periods when prices were unstable.
- 30.
Recall that the number of remaining miners is fixed from 2001, so an alternative series of growth rates from then on will have only a marginal impact on the evaluation of restructuring paths.
- 31.
Note that the sum is over the infinite horizon. This is conceptually proper if we view restructuring as a reallocation of labour that by default will be maintained indefinitely.
- 32.
Since the price and the wage are in units of current USD, the discount rate is a combination of the depreciation of the currency and time preference.
- 33.
See (4). Note that we are abstracting from the externality that miners impose on non-miners or vacant firms in the labour market. Whether this externality is overall positive or negative would depend on the labour market properties, in addition to those that we have assumed.
- 34.
The relatively favourable labour market outcome in Ostrava may be an exaggeration since we have assumed extra unemployment linked to, and in proportion to, mining unemployment in all periods in Jiu Valley, but not in Ostrava (see Sect. 3.2). If we also assume extra unemployment linked to mining in Ostrava, the resulting unemployment rate would be higher although this effect would be mitigated by the slower restructuring pace (see Footnote 25). Regardless, under any reasonable size of extra unemployment, a large difference in the labour market outcome between the regions would remain.
- 35.
The inefficiency measure for the actual path, under these alternative parameter values, is 5.6% in Ostrava and 26.5% in Jiu Valley. The small change in inefficiency from the benchmark in Jiu Valley is the result of two off-setting effects: the inefficiency of inaction in the early 1990s is reduced but the speed of layoffs in the late 1990s becomes too fast.
- 36.
More generally, there is an argument that the workers’ unions fight over the size of membership as a means of ensuring future rents. See Acemoglu and Robinson (2001) for a formal treatment of this.
- 37.
The government was widely perceived as unreformed ex-communists. The miners’ action itself significantly contributed to this perception (see Sect. 4.3). Had the miners and the government pursued early restructuring as envisioned here, the creditors may have been more receptive.
- 38.
See Footnote 11 for the sources of the labour market data used here. The employer paid about 38% of wages as social security tax. The employee paid about 25% of wages as income tax. So the total tax rate is (25%+38%)/138%.
- 39.
The severance payment was in the range of 12–20 months of the miner’s wage. We estimated the government spending on severance payments in 1997 by the fall in the number of miners from the previous year multiplied by 16 months of the miner’s wage net of tax.
- 40.
The figure 350 is reached after excluding the last seven months of layoff, when the layoff pace jumps above 700 a month, reflecting the jump in actual spending associated with the 1997 layoff.
- 41.
We conducted a sensitivity analysis on \( \bar A \), λ and ρ, as in Sect. 3. The gradual restructuring path varies little across different values. This is because the budget constraint (12) is the binding constraint, little affected by these parameters.
- 42.
A 1993–1994 survey conducted by the mining company CNH shows that a majority of employees believed that downsizing would occur, but they were much less concerned about its danger than that of inflation (see Krausz 1999).
- 43.
For a general argument for gradual reform under uncertainty, see Dewatripont and Roland (1995) among others.
- 44.
To be precise, given a gradual restructuring path, calculate the expected discounted sum of wages and the severance payment during the remainder of his work life conditional on being laid off with the timing of layoff randomly distributed according to the layoff schedule. This sum is smaller than the discounted sum of wages conditional on not being laid off by the stated range of amounts.
- 45.
There may have been hidden deals between the miners and the government (see Sect. 4.3 for some clues). The union leadership may have had a private agenda: Cozma apparently became wealthy as well as politically influential while the miners’ action was successful. The confrontations between the miners and the government in the past may have affected the expectation on the events in the 1990s (more on this below).
- 46.
The government may have chosen the compensation package rather than a more comprehensive labour conversion program because the compensation package minimized the resistance from the myopic miners. This is in line with the political aspect of restructuring, discussed in Sect. 4.3.
- 47.
The above discussion can be placed in the context of the “socialist patrimonialism” in Romanian society during the communist era (see Linden 1986).
- 48.
Looking beyond the two regions of our inquiry, it is not new that a labour union may fight a losing battle to the end. A good example is the 1984 British coal miners’ strike. The British miners went on a year-long strike against restructuring despite the offer of a generous severance payment. In the end the strike achieved nothing but lost pay and a ruined reputation for the miners. The miners, at least the union leadership, seem to have been driven by more than the narrow economic interests of the miners; perhaps they were driven by the ideology of socialist order or, more specifically, the involvement of the union in management decision making. The tripartite agreement, achieved ten years earlier through a series of strikes, indeed guaranteed something of this nature. The turning point was the attempt by the Thatcher government to nullify this arrangement. Under a threat to the status quo and under the memory of previous successful strikes, the miners may have been overly prepared for action, not unlike the Jiu Valley miners. See Edwards and Heery (1989) for an excellent study of labour relations leading up to that episode.
- 49.
This interpretation can be couched in an extension of the argument in Footnote 36: a government tries to affect the size of a union as a means of ensuring the benefit, or eliminating the harm, from the union.
- 50.
The epilogue of this sequence of events is the re-emergence of the conservative faction in the 2000 election. By then the main phase of mining restructuring had been completed.
References
Acemoglu, D. & Robinson, J. (2001). Inefficient redistribution. American Political Science Review, 95(3), 649–661.
Aghion, P. & Blanchard, O. (1994). On the speed of transition in central Europe. NBER Macroeconomics Annual, 9, 283–320.
Bruha, J., Ionascu, D., & Jeong, B. (2007). Organized labor and restructuring: Coal mining in the Czech Republic and Romania. Finance a uver, 57(5–6), 272–291.
Castanheira, M. & Roland, G. (2000). The optimal speed of transition: A general equilibrium analysis. International Economic Review, 41(1), 219–239.
Chiribuca, D., Comsa, M., Dincu, V., & Rotariu, T. (2000). The Impact of Economic Restructuring in Mono-industrial Areas. SOCO Project Paper n. 87, Institute for Human Sciences, Vienna, Austria.
Czech Mining Institute. (1993–2002). Mining Yearbook. Ostrava: Czech Republic.
Dewatripont, M. & Roland, G. (1995). The design of reform packages under uncertainty. American Political Science Review, 85(4), 1207–1223.
Edwards, C. & Heery, E. (1989). Management control and union power: A study of labour relations in coal-mining (p. 276). Oxford: Clarendon.
European Union. (1996). Clean coal strategy for central and eastern Europe: Draft status review. Brussels: Social Policy Analysis Group.
Galdon-Sanchez, J. & Schmitz, J. (2002). Competitive pressure and labor productivity: World iron-ore markets in the 1980’s. American Economic Review, 92(4), 1222–1235.
Kideckel, D. (2000). A new ‘cult of labor’: Stress and crisis among Romanian workers. Sociologie Romaneasca, 1, 142–161.
Krausz, S. (1999). Inertia of mentality as a break of transition: An example of attitude toward the downsizing of economic activities. In S. Krausz (Ed.), Sociology of transition. Petrosano: University of Petrosani.
Larionescu, M., Rughinis, C., & Radulescu, S. (1999). With the eyes of the miner: The reform of the Romanian mining sector (p. 275). Bucharest: Gnosis.
Linden, R. (1986). Socialist patrimonialism and the global economy: The case of Romania. International Organization, 40(2), 347–380.
Olson, M. (1982). The rise and decline of nations (p. 273). New Haven: Yale University Press.
Parente, S. & Prescott, E. (2000). Barriers to riches. Cambridge: MIT.
Putnam, R. (1992). Making democracy work: Civic traditions in modern Italy. Princeton: Princeton University Press.
The Economist. (1999). “Romania: All very sinister,” Edition of February 6, London.
Acknowledgement
We are grateful for funding from the Global Development Network. We thank, without implicating them, Ayse Imrohoroglu, Edward C. Prescott, James Schmitz and the anonymous referees, as well as seminar participants at CERGE-EI, the Global Development Network Workshop in Prague, the International Society for New Institutional Economics Conference in Boston, the European Association of Labor Economists Conference in Seville, the Universitat Pompeu Fabra in Barcelona, and Copenhagen Business School for helpful comments.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2010 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Bruha, J., Ionascu, D., Jeong, B. (2010). Organized Labour and Restructuring: Coal Mines in the Czech Republic and Romania. In: Caroleo, F., Pastore, F. (eds) The Labour Market Impact of the EU Enlargement. AIEL Series in Labour Economics. Physica-Verlag HD. https://doi.org/10.1007/978-3-7908-2164-2_3
Download citation
DOI: https://doi.org/10.1007/978-3-7908-2164-2_3
Published:
Publisher Name: Physica-Verlag HD
Print ISBN: 978-3-7908-2163-5
Online ISBN: 978-3-7908-2164-2
eBook Packages: Business and EconomicsEconomics and Finance (R0)