Chapter 3: improving labour market performance.
3.1 Despite some improvements in the labour market situation during the 1990s--with estimated unemployment (36) declining from its peak of 12% in 1997 to 8.7% in 2001 and employment increasing by 14% between 1993 and 2003, the high level of unemployment remains a major policy concern. France also has one of the lowest rates of utilization of labour potential among OECD countries (OECD 2004a.). This not only reduces living standards and growth potential but it also reduces government revenues and increases expenditures on the non-working population.
3.2 The low utilisation of labour potential is mainly caused by four factors:
* high structural unemployment,
* low labour market participation among the youth,
* early exit from the labour market by older workers,
* a relatively low number of annual hours worked per worker.
3.3 The main challenges therefore are to:
* reduce impediments to labour market adjustment such as unduly strict employment protection legislation (EPL) and high tax wedges;
* reduce or remove inappropriate incentives that lead to early withdrawal from the labour market for particular groups such as the young, the old and mothers of young children;
* ensure that social legislation does not unduly impinge on the ability of workers and employees to negotiate working hours;
* maximise cost efficiency in job placement services.
3.4 Past policies have addressed some of the challenges, sometimes successfully. Changes have mainly focused on active labour market policies and on lowering taxes for low income earners, leaving other labour market rigidities unchanged. Some of these policies are costly for the budget and there may be more cost efficient measures available. In addition, some measures went in the wrong direction, in particular the introduction of early retirement programmes, the increase in EPL from an already relatively high level and also the compulsory reduction in working hours. Although the tax wedge declined for low paid workers, it remains high for average and higher-skilled workers. In addition, benefit entitlements for some groups of unemployed increased and the implicit tax on continuing work for older workers, though it has fallen, remains very high. Hence, despite the reforms carried out in recent years, France has continued to lag behind many other OECD countries in implementing Jobs Strategy recommendations (Brandt et al. 2005). Therefore, it seems of crucial importance for France to further reduce barriers to employment by comprehensive reforms which address problems on both the demand and the supply side of the labour market.
3.5 This chapter argues that reform of the labour market needs to combine measures to restore incentives for firms to hire workers and for workers to stay as long as possible on the labour market, ensure that employers and employees have greater ability to negotiate working time, and improve the efficiency of services for employment so as to reduce long term unemployment. The main areas addressed here are employment protection legislation, constraints on working hours, wage formation, taxation of labour, the public employment service and policies that lead to withdrawal from the labour market.
Box 3.1 Summary of recommendations to improve labour market performance Reform employment protection legislation * Reduce the strictness of EPL and the difference between permanent and temporary contracts by introducing a single contract with lower protection that increases with the length of employment. Otherwise, or as a temporary measure, the cost of "economic" dismissal could be reduced and the use of temporary contracts eased * Reduce the cost of the obligation on firms to help their employees to find a new job * Suppress or further reform the Delalande contdbution--the penalty paid by firms for dismissing workers more than 50 years old Ensure that employers and employees have the ability to negotiate working hours Continue reducing the cost of unskilled and part-time labour * Limit future increases in the minimum wage and renounce the use of "coups de pouce" for a certain period * Ensure the stability of policies of cuts in social security contributions so as to enhance their efficiency by avoiding uncertainty generated by frequent changes Improve the efficiency of the public service for employment * Improve markedly coordination between the job placement agency (ANPE), the unemployment insurance administration (UNEDIC) and other agencies * Create a one stop shop service system for the unemployed * Improve monitoring of job search and make use of systematic evaluations of these policies to improve efficiency Continue reducing inactivity traps * Carefully evaluate the aim of family policies and avoid inappropriate disincentives to employment * Phase out early retirement programmes and increase older workers' obligations to look for a job.
The development of the labour market over the past decade
Despite recent improvements structural unemployment remains high while employment and participation are low
3.6 After deteriorating in the beginning of the 1990s, the labour market improved in the second half of the 1990s. Both actual and structural unemployment are below their mid-1990s peaks although actual unemployment increased again during the recent cyclical downturn (Figure 3.1). The employment rate increased in France by 4 percentage points after 1996 while the OECD average remained unchanged. In contrast with the earlier period, this growth came largely in the private sector (Figure 3.2). Employment also kept up quite well when GDP growth slowed in the early 2000s.
[FIGURES 3.1-3.2 OMITTED]
3.7 Up to the mid 1990s, growing unemployment had been largely at the expense of the least educated, whose share in employment declined steadily. These trends have been halted and reversed as unemployment among the least educated workers decreased markedly in France after 1992, while it increased in many other OECD countries (Figure 3.3). Youth unemployment has also fallen significantly since 1998.
[FIGURE 3.3 OMITTED]
3.8 Despite these improvements France continues to belong to the group of OECD countries with relatively poor labour market performance (Figure 3.4). Moreover, since 2002 long-term unemployment has increased more markedly than the OECD average, and more so than in many euro-area countries (Figure 3.5). Although young people are perhaps the most heavily affected by unemployment, the unemployment rate for prime age workers is also high. Employment rates are also well below the OECD average for older workers.
[FIGURES 3.4-3.5 OMITTED]
Working hours have declined significantly
3.9 Annual hours worked per capita are among the lowest in the OECD and this is the main source for the gap in GDP per capita and for its widening vis-a-vis the best-performing countries in the OECD (Figure 3.6). GDP per hour worked is relatively high in France, higher than in the United States. However, this good productivity performance is largely due to the exclusion of large parts of the unskilled labour force from employment in France (Artus and Cette 2004). This low utilisation of labour potential is mainly due to low hours worked per worker, in particular the low working-time of full-time workers although the low employment rate also contributed. (37) The same factors contributed to the decline of labour utilisation over time (Figure 3.7).
[FIGURES 3.6-3.7 OMITTED]
Labour market problems are mainly structural
3.10 The persistence of labour market problems through a number of cycles shows that they are structural and deeply rooted. Various factors may be at work, among them, institutions, certain aspects of wage bargaining, work incentives, and job search.
3.11 The rest of the chapter examines these various aspects of labour market practices and institutions. These institutions provide a high level of protection to workers, especially workers on permanent contract, which may lead to the perverse effect of curbing hiring. The legislation on hours is also quite restrictive, which could have a negative impact on output in the long-run. The chapter then analyses the system of wage setting. While wage formation is constrained for low paid workers because of the minimum wage (SMIC, salaire minimum interprofessionnel de croissance), it appears to be relatively decentralised and unconstrained for the rest of the wage distribution. Nevertheless, the high level of unemployment suggests that this wage flexibility is not sufficient to clear the market. One possible factor at work is the high level of social security contributions, despite some reductions. To assist policies to reduce rigidities and costs, efforts also need to be made to improve the efficiency of the public employment service and to stimulate labour supply, especially among certain groups.
The Labour Code, jurisprudence and administrative procedures provide high protection to workers Employment protection is rather strict and has strengthened
3.12 According to OECD indicators, EPL is relatively strict in France (Figure 3.8). This is largely due to highly restrictive dismissal procedures laid down in the Labour Code for permanent contracts--which is the standard form of contract allowed by the law, and is presumed when there is no written agreement. Temporary contracts, which ought to allow some flexibility around the constraints of the standard contract, are in turn highly constrained in their application (Box 3.2). Surprisingly, temporary contracts in the public administration are less constrained.
[FIGURE 3.8 OMITTED]
Box 3.2. Main characteristics of employment protection legislation Rights and obligations are defined in the Labour Code, which provides a minimum guarantee for workers' rights but collective bargaining agreements can achieve more favorable conditions for workers. There are two types of contracts, permanent or temporary (temporary contracts with the employer directly or with an agency), but the full time permanent contract is the reference one. Permanent contracts Provisions concerning permanent contracts mainly cover how to terminate them. There are three categories identified: dismissal, retirement or resignation. Dismissals can take two broad forms, dismissal for "personal" reasons or for an "economic" reason. Dismissal for "personal" reasons can take two forms, either for "real and serious" reasons, including refusal of a change in the labour contract or the inability to meet objectives set by the employer, or for "fault". In the latter case, the employer does not have to pay severance payment. A dismissal for economic reasons can be "individual" if it concerns a single worker or "collective" for two or more workers. In both cases, allowable justifications for such dismissals are strictly defined. It should not be inherent in the person but caused by economic difficulties or technological changes. Moreover, the 1995 jurisprudence has further narrowed the reasons for an economic dismissal (see Annex). It can be used only if it is necessary to preserve the competitiveness of the firm i.e., it cannot be used to improve competitiveness or profitability. Administrative procedures for layoffs for economic reasons are complex and long, more so when the layoff concerns many workers: * General provisions: the employers must respect a mandatory notice period and have interviews with the worker. Employers are obliged to help employees deal with changes in their jobs and finding another job. Firms also have to negotiate with the enterprise committee in the case of a collective dismissal. * A dismissal of more than 10 workers. An "employment preservation" plan (plan de sauvegarde de l'emploi), which includes measures such as encouraging the search for jobs outside the firm, creation of new activities, training programmes, etc, must be put in place. * Firms with more than 1 000 employees have to offer a "reclassification leave" which is a 4 to 9 month period before the employee is dismissed during which firms have to provide their employees with training and other help to find a job. Severance payments depend on the number of years spent in the firm and on the reason for the dismissal: * For an employee with less than 10 years of experience in the firm: 1/10 of monthly wage by year of seniority and 2/10 in case of dismissal for economic reasons. * For an employee with more than 10 years of experience in the firm: 1/10 of monthly wage by year of seniority for the first 10 years and 1/15 of monthly wage by year of seniority after 10 years. In case of a dismissal for economic reasons, the rates are respectively 2/10 and 2/15. Dismissed workers have priority in any re-hiring done by the firm. Employees can bring the case before a court (the "Prudhommes") if they contest the reason for the dismissal or the procedures followed. If the reason for the dismissal is not serious and justified according to the court, firms have to pay the worker an allowance equal to at least six months. Temporary contracts Temporary contracts cannot be used to fill a permanent job linked to permanent activity of the firm. Therefore, their use is restricted to specific situations: * to replace an employee on leave * temporary increase in firm activity * seasonal jobs or jobs in specific sectors with a lot of fluctuations in their activity. In the latter case, they are called "common use temporary contract" and there is a list of sectors which can use them. The framework of the contract is strictly set: * the duration of the contract must be written in the contract. It can be renewed once only. * maximum duration is 18 months (including renewal), 9 months in certain cases, and 24 months for very specific reasons After a dismissal for economic reasons, for a period of six months it is not possible to hire a worker on temporary contract for temporary increases in activity or seasonal tasks. A worker on a temporary contract cannot be paid less than a worker on a permanent contract in the same firm with equivalent skill and position. Firms have to pay the employee a premium of 10% of the gross monthly wage when a fixed term contract expires and is not transformed into a permanent one. Public sector The State is not subject to the same restrictions on short term contracts. In theory, temporary contracts can be used only in specific situations: when no civil servants have the skills needed to do the job (which is the case of computer scientists for instance) or for an occasional need. In practice, these contracts are easily used. They may be of three years and can be renewed indefinitely. The salary can be less than that of a civil servant doing the same work, and most other advantages enjoyed by civil servants are not granted. The use of temporary contracts in the public sector is not consistent with the 1999 European directive. The State has planned to restrict the use of these contracts and to transform some of the existing temporary contracts into permanent ones. Efforts in these directions have been very slow up to now.
3.13 Legal provisions concerning permanent contracts mainly cover how to terminate them. Severance payments are not particularly high in France, representing half a monthly wage for an employee with 5 years of seniority in case of a dismissal for "personal" reason and 1 monthly wage in case of a dismissal for "economic" reason. For an employee with 15 years of seniority, severance payments reach 1.3 monthly wage for a dismissal for "personal" reason and 2.7 monthly wage in case of a dismissal for "economic" reasons (Figure 3.9). However, the procedure to be followed in case of a dismissal for "economic" reasons is very complex and increases employers' costs considerably. Moreover, firms also have to help displaced employees to find new jobs. This is costly and likely to be particularly difficult when the finn itself is in difficulty.
[FIGURE 3.9 OMITTED]
3.14 France is one of the few OECD countries where EPL concerning permanent contracts has increased from the mid 1970s to 200238, due both to new legislation and to jurisprudence (see Annex 3.A.1). EPL has tightened in three areas:
* Procedures for economic dismissal have become more complex;
* Legal restrictions on economic dismissals have tightened;
* The burden of the obligation to help redundant workers to find new jobs has increased.
3.15 The previous government introduced the Social Modernisation Law in 2002, significantly tightening the constraints on dismissal of more than 10 employees. But in 2003 the new government suspended some of these provisions before introducing another law in 2004 which, while moderating some aspects of EPL, increased the obligation on employers to try to find alternative jobs for employees under threat of collective dismissal.
Firms are reluctant to dismiss workers on permanent contracts
3.16 The law permits "economic" dismissal only if it is necessary to preserve the competitiveness of the firm. Financial rationalisation by the management is not sufficient justification. Moreover, in 2002 the Social Modernisation Law added a provision requiring that the financial position of the group to which the firm belongs should be taken into account, which means that an economic dismissal is not legally justified if the group is healthy. These provisions prevent firms from undertaking practically any reorganisation to increase productivity that might ensure the survival or faster growth of the finn in the future, if it involves job losses. The assessment of whether the economic situation of the firm justifies economic dismissals is made by a judge. The number of cases brought before the courts is high compared with other OECD countries (Employment Outlook 2004a.). In a study relating to the year 1995, it was reported that employers lose 74% of these cases, compared with 48% in Canada, 51% in Italy, and 38% in the United Kingdom (Bertola et al., 2000).
3.17 As a result of this strict EPL, job termination does not often occur through dismissal for "economic" reasons (Figure 3.10). Dismissals for economic reasons increased only moderately despite the fall in profitability after 2001 (Figure 3.11). Labour market weakness in this period was reflected first in a big increase in the termination of temporary contracts and, secondly, through an increase in dismissal for other reasons including those for "personal" reason, which are considered less costly by firms than dismissals for "economic" reason. Indeed, in the case of a dismissal for "personal" reasons, an agreement is often signed between the employee and the employer that ensures that neither will contest the procedure of the dismissal, and severance payments are negotiated. The cost of such agreements suggests that the constraints of EPL are costly to finns. Kramarz and Michaud (2004) estimate that the average cost of dismissing a worker represents 14 months wages but is significantly higher for a collective dismissal. Although this kind of job termination is easier for finns, it has three negative aspects. First, it leads to very unequal treatment between employees, with some receiving very high severance payments while others do not (Cahuc and Kramarz 2004). Second, since it absolver finns of their obligations to help workers find a new job, the help that is presumed to have been given by firms is in fact never provided to these dismissed workers. Finally, a possible consequence of this kind of employment adjustment is that high productivity workers quit the firm while low productivity workers remain in the firm with negative implications for productivity.
[FIGURES 3.10-3.11 OMITTED]
Temporary contracts have developed strongly
3.18 Despite the constraints on their use, temporary contracts developed markedly during the 1990s (Figure 3.12). It is likely that some firms also manage to by-pass some aspects of the law, especially the one according to which it can be renewed only once (e.g., by creating another, similar, job). Thus, these contracts are more flexible and therefore attractive to firms. While they represented less than 5% of dependent employment in the mid 1980s, their share reached 14% by 2002, slightly above the OECD average. The contribution of temporary contracts to employment growth since the mid 1980s has been significant.
[FIGURES 3.12 OMITTED]
3.19 Temporary contracts in France are of very short duration, less than 5 months in most sectors (Table 3.1). The most frequent duration in other countries is between 6 months and a year (OECD 2002). The short duration does not seem to be a consequence of the limitation on the maximum duration of those contracts, which is 18 months, much longer that the average effective duration. It is more likely to be a consequence of the fact that the employer must pay for the entire duration of the contract even if the job is terminated before the date specified in the contract.
3.20 As a result of restrictions on fixed term contracts, they are mainly used for seasonal tasks with well known duration. They are less suitable when employers do not know precisely how long they need someone. For instance, they are appropriate neither for medium term projects nor for periods during which the recovery seems uncertain. Hence, temporary contracts often correspond to low skilled jobs of short duration with poor prospects for human capital accumulation. It is difficult for workers to obtain temporary contracts of good quality, not restricted to seasonal tasks that will help them to remain employed afterwards.
Labour market dualism has increased
3.21 Despite the fact that temporary contracts are often used for seasonal tasks, they provide flexibility to firms, helping them to rapidly adjust their workforce to changing conditions while the probability that a temporary contract is transformed into a permanent one is low (the probability being greater for workers employed directly, at 1 chance in 3, than for those under contract to an agency, for whom the probability is only 1 in 4). Such behaviour can also be observed in other OECD countries where strict employment protection of permanent workers tends to increase the incidence of temporary work (OECD, 2004a.). As a result, labour market dualism is increasing with the majority of workers benefiting from permanent contracts and high employment protection, while a growing number alternate between unemployment and short temporary contracts.
Overall strictness of employment protection legislation should be eased
3.22 The effect of employment protection on labour markets is a controversial subject and the literature highlights positive and negative effects on labour market performance (OECD, 2004a.). However, there seems to be a consensus on two points. First, while strong EPL and low unemployment initially coexisted, circumstances have changed and the interaction between economic shocks and EPL can explain part of the increase in unemployment and unemployment persistence (Blanchard and Wolfers, 2000). Second, EPL does not affect groups in the same manner and is likely to have negative effects on youth, prime age women and older workers (OECD, 2004a.). Some arguments seem to be relevant for France:
* The purpose of EPL is to protect existing jobs. At the same time, it increases the present value of the cost of hiring. Moreover, for low skilled workers and young workers with low education levels, the minimum wage already makes this high, despite the reductions in social contribution rates on low salaries. Taken together, these measures significantly lower the chances of these workers finding a job.
* The impact on wages depends on opposing factors. On the one hand, EPL might increase wages by increasing insiders' bargaining power, but, on the other hand, it could reduce the reservation wage because it reduces risk. Moreover, by reducing the hiring rate, it should decrease the reservation wage for new entrant workers. However, as the minimum wage puts a floor on the wages of new entrants, the effect on reservation wages is reduced or eliminated. As a result, EPL on permanent contracts is likely to reduce wage flexibility.
* Theory suggests that differences in the strictness of EPL on temporary and permanent jobs produce high turnover in temporary jobs, since high firing costs for permanent contracts induce firms to use temporary contracts in sequence rather than converting them to regular contracts (Blanchard and Landier, 2002, Cahuc and Postel-Vinay, 2002, OECD, 2004). As a result, easing the use of temporary contracts would foster both hiring and firing with unclear implications for overall unemployment. This might be the case in France where one-year and two-year mobility of workers on temporary contracts towards permanent ones is low relative to other European countries (OECD, 2002). Nevertheless, temporary contracts give some low productive workers the opportunity (that they would not have had otherwise) to enter the labour market and to acquire some skills that may improve their employability.
3.23 The objective of a reform in EPL is to increase the likelihood that unemployed workers find a job and stay employed, avoiding long term unemployment. A possible reform to increase employment could include:
* easing the use of temporary contracts so that they can be used for permanent activity in risky sectors, which should incite firms to take more risk and to create jobs that, if matched with the right workers, would ensure human capital accumulation and increase productivity;
* reducing the cost of dismissal on permanent contracts in order to increase employers' willingness to hire on permanent contract or to transform temporary contracts into permanent ones.
3.24 A number of reports have made suggestions on possible ways to make these changes:
* The de Virville report (2004) suggests relaxing the legislation on temporary contracts by enlarging the access to "common use temporary contract" (contrat d'usage), which already exist for specific sectors and whose use has already been extended, through the jurisprudence, to other sectors (see Annex);
* Blanchard and Yirole (2003) and Cahuc (2003) suggest reforming EPL on permanent contracts by reducing the burden of administrative procedures, lowering the cost to firms of helping dismissed workers to find another job, and limiting the judge's power. They suggest introducing, at the same time, an experience rating system under which employers' contributions are made an increasing function of the frequency with which they dismiss workers;
* Cahuc and Kramarz (2004) suggest abolishing the difference between permanent and temporary contracts by creating a single contract, with severance payments increasing with the length of service. Another suggestion is that firms pay tax to public authorities that will finance actions to help dismissed workers to find a job. In exchange, the procedures and the allowable reasons for "economic" dismissal would be smoothed.
3.25 The creation of a single contract is likely to be the most efficient solution, as it should not only increase effective labour demand but also reduce labour market duality and improve fairness between workers. Since implementing such changes may be politically difficult, gradual reforms could be introduced that would at the same time ease the use of temporary contracts and reduce EPL on permanent contracts. But some suggestions, such as designing further special types of employment contracts, are likely to increase administrative cost and reduce transparency, without a significant impact on reducing duality. In any case, the authorities should consider propositions to relieve firms of the obligation to help dismissed workers find a new job, making them contribute to financing the public employment service instead.
Legislation has significantly reduced hours worked
3.26 As mentioned above, working hours per worker are relatively low in France. Annual hours worked per employee are now 8% below the EU average and around 20% lower than in the United States, Japan, Australia or New Zealand (Figure 3.6). This is not caused by a particularly high incidence of parttime work but rather by the low annual working time of full-time workers (Table 3.2). Largely as a result of past and recent policy measures, the standard working time per week and per year is now significantly lower than in most other industrial countries.
3.27 Over the past decades France has legislated significant cuts in standard working hours per week which has also reduced the actual working time per year. In 1982, legal working time (i.e., standard hours in employment contracts, above which an overtime premium must be paid) was reduced from 40 to 39 hours per week. In 1996, the loi de Robien offered reductions in employers' social contribution rates to create new jobs or preserve existing ones through work-sharing by reducing working time, and in 1998 the first loi Aubry announced the reduction of the legal working week from 39 to 35 hours with effect from January 2000 for firms with more than 20 employees and from January 2002 for others. Firms were encouraged to reduce hours worked before 2000 through cuts in social security contributions that were made conditional on a 10% decrease in hours worked and the creation of 6% more jobs. The second loi Aubry (2002) confirmed the new legal working time and introduced new cuts in social contributions for firms reducing hours worked. This time, these cuts did not depend on the amount of the reduction in working time or on job creation.
3.28 The 35 hours legislation (39) prompted an increase in pay (for overtime) after 35 hours worked and decreased maximum allowable overtime (Table 3.3). It also introduced a guarantee for the monthly earnings of minimum wage earners in order to prevent a fall in their real income as a result of working fewer hours (see below). (40)
3.29 Other aspects of implementing the cuts in working hours were left to collective bargaining between finns and trade unions. These negotiations introduced greater working time flexibility so that finns can ask for longer or shorter hours per week according to their needs provided that the average annual work week is 35 hours. This increases firms' ability to respond to seasonal and cyclical fluctuations of production, and has been favourably appraised by employers (DARES, 2003). Negotiations also led to wage moderation for two or three years (Pham, 2002). Wage moderation, productivity gains through organisational changes and cuts in social contribution were the three pillars of the measures that were designed to limit the increases in firms' costs that the working hours reduction would have caused.
3.30 The driving force for the reduction in working hours in France (and some other European countries) was in particular to create more jobs by "work sharing" (on the assumption that insufficient aggregate demand rather than high costs was at the origin of unemployment) rather than to provide workers with more leisure by distributing part of productivity gains as more free time. In France it has also been justified as a way to improve the efficiency of cuts in social security contributions on the argument that firms had to hire workers in order to avoid a reduction in output (DARES, 2003). Many employers opposed this legislation as they feared an increase in labour costs although the measure was accompanied by fiscal incentives and wage moderation (42). The policy of shortening working hours came under additional attack when it was extended to smaller finns and also when hospitals claimed that the 35-hour week had led to severe staff shortages.
3.31 This paved the way for reconsidering the policy of mandatory working time reductions and introducing instead more flexible arrangements, in particular by the loi Fillon (OECD, 2003), by the collective bargaining reform (see Annex 3A.2), that introduced the possibility for enterprise agreements on hours worked to be less favourable to employees than sectoral ones, and by the new arrangements of 2003 and 2005 (Table 3.3). However, except for some firms facing specific problems where hours have been increased and wage moderation continued (Bosch, Doux, Seb); few firms have so far used the possibility to increase hours (Bilan de la negociation collective, 2004). While the relaxation measures taken since 2003 have cut firms' overtime costs sharply (especially between the 36th and the 39th hours), the situation probably stems from the recent weak economic outlook. Another reason could be that employers fear that the advantages of increased flexibility and wage moderation, which had come in exchange of reduction in hours worked and after long negotiations, could be lost again.
3.32 It is hard to evaluate the overall impact of the reduction in working time, the associated reduction in social security contributions, related negotiations at firm level on working practices and wage increases that were moderated for a certain period. It is fairly clear that the simple imposition of reduced working hours, with a compulsory overtime premium and a binding ceiling on total hours worked, would have been costly for the economy. A necessary condition for the 35 hours week legislation to have had a positive impact on employment was that productivity gains should be higher than the increase in the unit labour cost.
* Some estimates have evaluated the firm-level impact of the 35 hours week legislation by comparing job creation in firms that reduced working hours with that in similar firms that maintained the 39 hours. This methodology relies on the possibility of comparing behaviour of similar finns, which is very difficult task and was not possible once the legislation was fully implemented since only small firms with specific characteristics maintained a 39 hour week. Such estimates conclude that the 35 hour week legislation increased employment by 6% for firms eligible for incentive provisions (the loi Robien and the first loi Aubry) and by roughly 3% for other firms shifting to a 35 hour week (the second loi Aubry) (Jugnot 2002), meaning that productivity gains, wage moderation and cuts in social contributions were able to offset cost increases for businesses. This result is consistent with survey studies in which firms consider that 6% of employment was generated by the 35 hours week legislation (Bunel et al., 2002).
* Empirical studies all agree on the fact that the legislation was differently appraised among firms. While large firms say they are happy with the changes in working practices, notably on the flexibility on week-to-week or month-to-month allocation of working time, that they were able to negotiate in exchange for an early move to the 35 hours week, smaller firms were not able to undertake re-organisation and thus were more affected by the increase in labour cost (and even if the 35 hour week was introduced for SMEs later on).
* The overall impact on employment of the 35 hour week legislation is difficult to estimate. Assuming that the legislation did not affect production negatively in the short term, and adding the different firm-level job creations, Jugnot (2002) finds that the overall impact is a 310 000 net creation of jobs in the short term. This estimate also assumes that firms that maintained the 39 hour work week were not affected by the legislation, which may not be true since the SMIC grew more rapidly in the first few years of implementation of the legislation. This was because the SMIC is linked to the average hourly wage, which accelerated as working time reductions concerned more and more workers, because monthly wages were very often kept constant when working time was reduced.
* The long-term impact of the 35 hours legislation is also very uncertain (Pisani-Ferry, 2003). No studies have taken into account the cost of financing the 35 hours week legislation that, in the end, is very likely to require higher taxes. This cost includes the cuts in social contributions that were made during the introduction of the legislation and also those that came after 2002, during the period of harmonisation of the different minimum wages, to offset the increase in wages that this harmonisation caused. Furthermore, while the increased flexibility has further increased growth of labour productivity per hour, this was not enough to prevent a clear deceleration of growth of productivity per employee (Figure 3.13). This is likely to have reduced growth of potential output and (material) living standards. Finally, no studies have taken into account the fact that reducing working hours may have led to an underutilisation of the capital stock that could also have a negative impact on growth.
[FIGURE 3.13 OMITTED]
Wage setting is constrained for wages close to the minimum wage but is relatively decentralized otherwise
The minimum wage has increased as a consequence of the 35 hours legislation
3.33 The Labour Code specifies a legal minimum wage. Before 2002, there was a single minimum wage, the SMIC, indexed to inflation and half the increase in the purchasing power of the average wage, with occasional discretionary increases. When the legal workweek was reduced from 39 hours to 35 in any particular firm, the monthly earnings of minimum wage earners were maintained at the level applying in the year in which the firm reduced hours. As a result, the 35 hours workweek legislation created several hourly minimum wages, according to when a worker's employer signed a 35 hours workweek agreement (OECD, 2003). By July 2005, these different minimum wages will be harmonised and there will be a single SMIC, as before. During the period of harmonisation (from July 2002 to July 2005), the average real minimum wage (excluding employer contributions) will have increased by 6.5%. The way the SMIC will evolve after 2005 has not yet been decided, while this year the SMIC is to revert to its usual method of indexation.
Wage formation appears to occur mainly at firm level
3.34 Collective bargaining takes place at various levels in France: national, inter-professional, sector, section of industry and firm level. Five unions are officially considered to be representative of employees, having an "indisputable presumption" that automatically involves them in bargaining (CFDT, CFTC, CFE-CGC, CGT, CGT-FO). (43) Other unions have to prove their representative character at the level of the bargaining unit. There are no constraints on the side of employers. Bargaining covers various issues apart from wages. Sector-level collective agreements are very often extended to all employees of the sector. The OECD index of legal extension of sector-level collective agreements is thus high for France (Brandt et al. 2005).
3.35 In practice, however, enterprise level bargaining has developed strongly and this tendency has been reinforced with the 35 hour week legislation; the relevance of having branch level bargaining is regularly questioned and the number of branch agreements is decreasing. One of the reasons is that, because of the large discretionary increases in the minimum wage, minima negotiated at the industry level have been overtaken by the SMIC. In 2003, the sector level minimum was below the national one in 78% of the branches. Furthermore, almost all agreements on wages and part of the extensions in fact aim to raise the sector-level minimum wage up to the level of the national one. As a result, intermediate level negotiation on wages does not seem to play an important role in wage setting. There are reasons for expecting that this move towards more decentralised wage bargaining should reduce structural unemployment (OECD 2004a.).
Nominal wages show downward flexibility and average wage growth has been moderate
3.36 Empirical studies also conclude that wages seem to be mainly determined at firm-level. Theoretically, firm-level wage setting should allow firms to adjust workers' wages to changes in their productivity. As it is difficult to measure individual productivity, studies have first measured flexibility by the existence of negative adjustments of wages and the absence of a spike at zero in the frequency distribution of wage changes. According to this definition, employee-level data display wage flexibility (Goux, 1997, Biscourp and Fourcade, 2003). Each year, one quarter of workers who do not change employer experience a decrease in their annual wage and 20% of them a decrease in their hourly wage. This result has been shown on different sets of data from 1991 to 2000. Since the Labour Code tends to protect workers from decreases in their basic wages (Bonnechere, 2002), adjustments come from decreases in variable pay schemes, as offsets to improvements of working conditions or changes of jobs inside the firm. When flexibility is measured as the reaction of wages to positive and negative shocks to firm activity, flexibility appears to be less "perfect" (Biscourp and Fourcarde, 2003). Wages seem to react more to a positive shock than to a negative one. Part of the explanation must be that the minimum wage constrains downwards adjustments.
3.37 Slowdown of average real wage growth, that can be observed at the aggregate level from 1985, with wages accounting for two thirds of the reduction in labour cost growth between the periods 1970-85 and 1986-2003 (Table 3.4).
3.38 Real wage growth has remained moderate in recent years, without falling below that of productivity, and as a result, wages have on the whole kept stable as a percentage of corporate value added since the mid-1990s.
Labour taxes have been cut for low-paid workers but labour cost at the level of the SMIC remains high
3.39 Taxes on labour are high in France compared to other countries (see Chapter 2). Beginning in 1993, the government introduced rebates on social security contributions for wages close to the minimum wage. Both the range of wages concerned by these rebates and the size of the rebates has increased with time. Between 1998 and 2002, the rebates were larger for firms that decided to reduce hours worked. Since 2003, the aim has been to offset the increase in the minimum wage generated by the harmonisation of the various minimum wages created by the 35 hours week legislation. With those rebates, taxes on labour now represent 46% of the net wage for a worker paid at the SMIC rising to 80% of the net wage for a worker on 1.6 times the minimum wage (Figure 3.14). As a result of these measures the tax wedge has strongly decreased for low paid workers but has remained unchanged for an average productivity worker (Brandt et al. 2005). Although it is difficult to estimate precisely the number of jobs created by such a policy, employment of low skilled workers has strongly improved and total net job creation has been strong since the introduction of these policies. Econometric estimates of the effect of the reduction of social contribution rates between 1993 and 1996 vary from 100 000 to 500 000 jobs.
[FIGURE 3.14 OMITTED]
3.40 Despite these tax reductions, unemployment of low-skilled workers is still high while at the same time, the trend towards reducing the labour cost of low-paid workers has been halted since the late 1990s. Given the limited fiscal room for manoeuvre, the only way to further lower low-skilled unemployment is likely to be to reduce the SMIC relative to the average wage, e.g., by blocking any real increase in the SMIC in the coming years.
3.41 Up to now, the framework for social security contribution rebates has been unstable. The wage threshold under which social contributions are reduced has changed almost every year, starting from 1.2 times the SMIC in 1993 to 1.6 as from 2005. The uncertainty both on the durability of the rebates and on their shape may have limited the positive impact on labour demand. Hence, the government should pursue more stable policy to ensure that firms' anticipations expectations can be more firmly anchored. Moreover, when changes in the range of wages concerned by these policies are made, the trade-off between targeting the policy on specific groups and increasing steeply marginal tax rates over the range where the rebates are phased out should carefully be taken into account.
The unemployment benefit system has been reformed but inactivity traps remain
3.42 While reducing labour cost further for low paid workers and easing EPL would stimulate labour demand, efforts are also needed on the side of labour supply in particular by ensuring that public employment services provide efficient counselling and monitoring of the unemployed and that inappropriate incentives to withdraw from the labour market are removed.
The organisation of the public employment service is complex
3.43 With more than 3% of GDP, France is one of the OECD countries with the highest spending on labour market programmes. Of these spending, 1.25% is used for active labour market policies. However, despite this spending, long-term unemployment remains high. Many reports have questioned the efficiency of the French public employment service.
3.44 In France, various authorities are responsible for the evaluation, placement and payment of job seekers as has been highlighted in the Marimbert report (2004). The three main authorities are:
* The Assedic (federated in the UNEDIC, union nationale pour l'emploi dans l'industrie et le commerce) handle the administration of the unemployment insurance and, since 1996 of the registration of any unemployed (before, it was the agence nationale pour l'emploi, ANPE). They are private associations administered by social partners from both the employer and employee sides.
* The ANPE, a public institution, until recently had a legal monopoly on job placement for the unemployed (the Act of 18 January 2005 abolished this monopoly). It is made up of several hundred agencies and secondary offices whose purpose is to gather job offers and requests and to provide advice to the unemployed in their job search. However, since 1986, other institutions, if recognised by the ANPE, may also offer placement services.
* The regional and departmental Directions du travail, de l'emploi et de la formation professionnelle are part of the Ministry of Labour and are in part responsible for the implementation of government labour market policies.
3.45 In addition to those three main actors, a national agency (the association nationale pour la formation professionnelle des adultes (AFPA)) is in charge of adult training and another (the agence pour l'emploi des cadres (APEC)) looks at matching job supply and demand for white collar workers. Moreover, the ANPE out-sources part of its activity to various external operators (Balmary et al., 2004). As a consequence of this complex organisation, job seekers may have to deal with several organisations. For instance, the financing of some special programmes for unemployed may depend on all three main authorities.
3.46 This complexity generates several problems:
* Resources are wasted by the multiplication of administrative structures;
* Job seekers can be discouraged by the complexity of the system;
* The monitoring of job seekers is inadequate (Marimbert, 2004, see below);
* Evaluation of the performance of authorities in charge of the placement is made difficult because of the number of entities in charge of it. As a result, evaluation is very rare (Balmary report, 2004).
3.47 Improvements have been made with the institution of the plan d'aide au retour h l'emploi (PARE, plan to help to bring people back to work) in 2001 that bring the ANPE and the Assedic closer. Despite this improvement, the system is still handicapped by the multiplicity of agencies. A further bringing together of those various services is still intended as part of the Social Cohesion Plan (Box 3.5). The proposition to create "houses for employment", places where all the agencies of labour market policy should be gathered, goes in the right direction. However, given the limited number of those houses scheduled, it seems unlikely that they could cover all the unemployed. The important point is that those houses be made one-stop-shops for the unemployed, gathering all the actors of labour market policy. If they only add more complexity to the system, the objective will be completely missed. The question is whether effective one stop shop services are possible without merging the institutions.
Box 3.5. Policies to improve employment of older workers The employment rate of older workers is among the lowest of OECD countries. Part of the explanation of this poor performance comes from the development of early retirement programmes in the 80s that were introduced hoping to reduce unemployment and facilitate restructuring. Although the government has ended most of these policies, problems on the labour market for older workers remain. These issues are discussed in depth in OECD (2005). The main points and main recommendations of this study are the following. Improving the incentives to continue working * Early retirement programmes have been tightened since 1990 for programmes provided by the State and since 2000 for programmes provided by the social partners but programmes provided by firms have also emerged and seem to have developed markedly. Remaining fiscal incentives for firms and employees to develop and to accept firms' early retirement programmes should be suppressed. Negotiations between the social partners should take remedial action to alleviate strenuous working conditions, giving no encouragement to early retirement. * Mutual obligations to job seekers and to the public service for employment should be gradually reestablished. Reform the exemption form looking for work that is available almost automatically to unemployed people after the age of 57 and half and to many after the age of 55. * Adjust the retirement age according to demographic trends. The average effective retirement age should rise gradually in parallel with the increase in contribution periods for a full pension, in line with increased expectancy--when the average effective retirement age is well over 60, the statutory minimum age and the compulsory retirement age should also be raised in line with life expectancy gains. Promote possibilities for benefiting from the bonus, the premium for people continuing to work after reaching the full pension entitlement, which is up to now limited to 65 years old. Implement good practices with regard to older public sector workers by reconsidering the very early retirement ages that exist for certain categories. Stimulating labour demand for older workers * Adding to general EPL, there is an additional penalty for firms dismissing workers older than 50 years (the Delalande contribution). Although the number of cases of exemption has increased, this penalty creates an entry barrier and reduces inter-firm mobility by increasing the cost of older workers. Moreover, there is no empirical evidence that this penalty decreases the number of dismissed older workers. Hence, it should be, either removed or reformed so that it really becomes an experience rating system. * Age profiles of earnings are very steep in France, which has a negative impact on labour demand for older workers, especially for low productive ones. Collective bargaining should consider linking wage evolution to skills rather than to age. Promoting the employability of older workers * On the job training opportunities are few in France compared with northern Europe countries. The 2003 inter-professional agreement should be implemented as soon as possible. On the job training at every stage of the career needs to be encouraged. Helping older worker to find a job and to continue working * Although specific programmes already exist for more than 50 years old unemployed, employment policies have mainly concentrate on the youth. PES initiatives to help job seekers over 50 back into work should be stopped up. * Make gradual retirement attractive and accessible to all. Support financially the switch to part-time work, though only for older workers on low income.
3.48 The fact that a large number of various programmes give young and low-skilled workers the opportunity to find a job by reducing the labour cost paid by firms (OECD, 2003) is another source of complexity and inefficiency of the Public Employment Service (SPE). The government plan to consolidate a number of these specific programmes to improve their performance is to be welcomed. The fact that more resources will be allocated to specific programmes in the public sector is more questionable since experience with this kind of jobs has been unconvincing in terms of providing people with lasting employment.
Box 3.3. The Social Cohesion plan The social cohesion plan is an ambitious plan that aims at reducing both structural unemployment and social exclusion. Provisions concern employment, housing and providing equal opportunities. The main measures regarding employment are the following. Public employment services * Opening of placement services to competition, which in fact already exists in practice, with a number of obligations (exemption from payment for the unemployed, free access, non discrimination). In return, the ANPE is allowed to make firms pay for its services. * Bringing together the State, the ANPE and the UNEDIC through a 3 year convention. In 2006, each unemployed should have a single file to which every organization should have access. * Creation of 300 "houses for employment" with a mission to provide access to the services of all of the agencies involved. * Improvement of the services offered to the unemployed will go with an obligation of active job search and an increase in monitoring. The unemployed have to participate in any action proposed by an agency. Unemployment benefits are to be suppressed if the unemployed refuse job offers that match their skills are appropriately paid and that do not lead to mobility inconsistent with their family situation. The possibility of reducing the unemployment benefit is introduced. Active labour market policies Specific programmes have been simplified leading to a halving of the numbers of programmes (from 14 to 7). Prorammes de end on the type of employer (public or private) and on the (recipient of a minimum subsistence income or not): * Recipient of a minimum subsistence income--employer in the private sector: minimum activity income (RMA). This is a part time job contract for six months (which can be renewed up to 36 months), during which training is provided. It is paid the hourly minimum wage but firms only pay the difference between the part-time minimum wage and the RMI. * Recipient of a minimum subsistence income--employer in the public sector: "contrat d'avenir". Same as the RMA except that jobs are in the public sector, on two-year contracts. In addition, employers are required to take support measures. * Other people facing substantial difficulties to find jobs (older workers, long term unemployed, disabled)--employer in the private sector: "Contrat Initiative Emploi" (CIE). Temporary or permanent contract with State help and with the obligation for the firm to provide training. * Other people facing difficulties to find jobs (older workers, long term unemployed, disabled)--employer in the public sector: "Contrat d'Accompagnement dans I'Emploi" (CAE). Same as the CIE except that jobs are in the public sector. Some of the other specific programmes (mainly those for the youth) remain: * "Contrat de professionnalisation": on the job training programmes that should lead skill acquisition. Workers can be paid below the minimum wage if they are less than 26 years old. Firms benefit from cuts in social security contribution. * "Contrat jeune en enterprise": part-time or full-time permanent contract in the private sector for low skilled youth between 16 and 25 years old. Firms benefit from cuts in social security contribution. * "Contrat emplois jeunes": programmes for youth in the public sector. New entries were stopped in 2004. These programmes should gradually disappear. Other Help to firm creation by unemployed. Cost of the employment part of the plan Million of euro constant, 2004 prices 2005 2006 2007 2008 2009 Houses for employment 120 405 530 375 300 Action for youth 235 452 525 513 468 Programmes for long- term unemployed 423 1 199 1 365 1 365 1 200 Development of on-the-job training programmes (apprenticeship) 169 297 402 512 629 Firm creation by unemployed 114 154 193 220 247 Total 1 061 2 507 3 015 2 985 2 844 Total, % GDP 0.1 0.2 0.2 0.2 0.2 Source: Social Cohesion Plan
Job seekers follow-up has increased but control is low
3.49 The unemployment benefit system was reformed in 2002. Conditions that need to be fulfilled to receive unemployment benefit, which depend on the length of affiliation and on the recipient's age, are now more restrictive but still relatively generous. The maximum duration has fallen from 60 months to 42 months. The minimum length of affiliation that gives access to the benefit has increased from 4 months to 6 months. Unemployed people who no longer receive unemployment benefit can receive another benefit for two years; this benefit, the ASS, (allocation de solidarite sprcifique), is designed to support the long term unemployed and is lower than unemployment insurance benefits.
3.50 Policy on monitoring unemployed job seekers was reformed in July 2001. On the one hand, unemployment benefits no longer decreased with duration of unemployment. On the other hand, individual follow-up has been made systematic, and more services are offered to the unemployed. The aim is to encourage active job search and increase its efficiency. A personalised action plan (part of the PARE) is now offered to any newly unemployed person, in the form of a contract between the unemployed and the unemployment institutions (Assedic and the ANPE). The former commit themselves to look for a job while institutions pay them the unemployment benefit and help them to find a job. An interview is now compulsory for any newly-registered unemployed person and recurs at least every six months. The purpose of the interview is to decide the degree of assistance and the type of services that the unemployed need.
3.51 At the moment, little is known about the impact of the personalised action plan (PAP). One study by Crepon et al. (2004) has attempted to evaluate its impact. The authors focus on the impact of specific support that concerns only 17% of the unemployed. This reinforced support is made up of four types of help: basic skill assessment, project assessment which is a deeper skill assessment, job search support and project support for people who have to change profession. The author's results depend strongly on the statistical technique used but are consistent with theoretical results. With a statistical correction for unobserved heterogeneity, only job-search support significantly reduces unemployment duration and increases the exit rate from unemployment to employment. Project assessment, job-search support and project support are found to decrease the exit rate from employment to unemployment. Hence, supporting the unemployed improves the quality of jobs found by the unemployed and thus decreases the probability of their becoming unemployed once again. Since the PAP is available only to a minority of the unemployed its macro-economic impact would, however, appear to be small.
3.52 Although the monitoring of job search per se has not been shown to decrease unemployment duration (Fougere, 2000), an increase in the monitoring conditional on more specific follow-up and help to the unemployed is likely to have a more positive effect on job search. The report of the Cour des Comptes (2003) has pointed to the lack of monitoring of job seekers as being also the result of the complexity of the system. In contrast with other countries, the monitoring of job searchers was not increased on the introduction of the PAP. The proposal in the "social cohesion plan" (see Box 3.5) to increase job search obligations and the introduction of the possibility of reducing the unemployment benefit (44) may address these deficiencies to some extent and could thus increase exit from unemployment into jobs when these obligations are not fulfilled.
Inactivity traps remain for certain groups
3.53 Relatively generous transfers including a relatively high replacement rate for the unemployed, a long duration for receiving unemployment benefits, a minimum subsistence revenue (RMI), family policies, and various local transfers (Anne and l'Horty, 2002) may discourage job search (Figure 3.15, OECD 2004b.). So as to provide incentives to take a job, the government introduced the possibility of taking a job while at the same time keeping the rights to some social benefits. Housing allowances and local residence taxes were reformed to alleviate aspects that could discourage people from returning to work. The government also introduced an in-work benefit in 2001 (the employment premium, in the form of a tax credit), which has been increased and extended in order to make part-time jobs pay. Finally, housing allowances and the local residence tax (taxe d'habitation) have been reformed to reduce disincentives to taking jobs.
[FIGURE 3.15 OMITTED]
3.54 Despite the introduction of these in-work benefits and other changes, there is still little incentive for one person of a household with two children to take a job paid at the minimum wage (Anne and l'Horty, 2002), because marginal effective tax rates are high. Certain family policies also give low paid women with children an incentive to withdraw from the labour market. When the allowance of "free choice of activity" (Box 3.6) was broadened to women with two children in 1994, the employment rate of women with two children (with one child younger than 3 years old) immediately fell by 11.2 percentage points (Piketty, 1998). The marginal effective tax rate is also high for part-time work at the minimum wage, even for a single person.
Box 3.4. Policies that discourage female labour force participation Some aspects of the family policy and of the tax system discourage female labour market participation (see Artus and Cette (2004)). Family policies Family policies are very complex in France but some of the measures have been unified very recently. Main measures are: * The PAJE (prestation d'accueil du jeune enfant) that aims at giving women "free choice of activity". The Paje lasts 6 months for a woman with one child and three years for a woman with more than one child. So as to be eligible, a woman must have worked two years during the last two years if she has one child, two years during the last four years if she has two children and 2 years during the last 5 years if she has more than 2 children. The amount is 512 [euro] per month if the mother does not work, 390 [euro] if she works on a part-time job lower than 50% and 295 [euro] for a part time job between 50% and 80%. * Assistance to lone parents (API). This allowance depends on the income of the person. Its maximum is 542 for a person with one child and 722 [euro] for a person with more than two children. This allowance lasts 3 years. If the person takes a job, part of the allowance can continue to be received. Tax system In a family, income tax depends on the income of the household, not of the individual. As a result, for a couple whose incomes are very different, the marginal tax rate for the person with the lowest income can be much higher than for a single person on the same income. Taking into account the cost of child minding, hardship of work or loss of leisure, the reservation wage can be higher than the gain of taking a job.
3.55 Where the unemployment rate is high and the minimum wage is binding, so that employment may depend more on labour demand than on labour supply, the gains from stimulating labour supply are uncertain. However, some analysis shows that part of the unemployment and low activity rate of certain groups, especially women with young children, may be explained by low incentives to take a job (Piketty, 1998, Laroque and Salanie, 2000). Since mainly low-skilled women are concerned, the risk of falling into a long-term inactivity trap is high. It may be that this is necessary if the aims of family policy are to be effectively pursued. It is not clear, however, that these aims have been carefully specified (for example, analysis of the impact of these policies on the children concerned is lacking), or the economic consequences of the policies have been properly assessed and considered.
3.56 Older workers are another major group for which the low employment rate is partly explained by low incentives to take a job (Box 3.7). Early retirement programmes have been developed in a context of high unemployment in a largely mistaken attempt to facilitate young workers' entry into the labour market. As a result, the employment rate of older workers plummeted between 1980 and 1985. Most early retirement schemes are being phased out. However, there is evidence that those schemes are effectively being replaced by specific schemes for workers who occupied physically demanding or stressful jobs.
The labour market needs a global reform
3.57 To summarise, specific characteristics of the French labour market tends to produce high structural unemployment and is particularly harmful for specific groups of workers.
* The high level of unemployment among prime age workers is the result of strict EPL and the high tax wedge that discourage firms' labour demand and of the poor efficiency of the public service for employment that increases the risk of becoming long-term unemployed.
* Young and low skilled workers are particularly affected by EPL (OECD, 2004a.) whose negative impacts on labour demand are reinforced by the relatively high labour cost at the level of the SMIC. These workers would need a very efficient public service for employment, capable of counselling them towards specific programmes that will at the same time improve their skill and provide them with a valuable first job experience.
* Low skilled women with children also suffer from restrictive EPL and high labour cost at the level of the SMIC. Moreover, incentives to withdraw from the labour market when their children are young are strong, with very negative consequences on their future employability.
* Older workers have also strong incentives to early withdrawal from the labour market. Moreover, EPL is particularly strong for older worker due to an additional penalty for firms dismissing workers older than 50 years (the Delalande contribution), even if the penalty was relaxed in 2003 (45). Research has shown that the penalty has had no positive effect on the employment of older workers (Behaghel et al. 2004, Bommier, et al. 2003). Hence, it creates an entry barrier and reduces inter-firm mobility by increasing older workers cost, which is already high because wages developments are strongly linked to age. Policies to reduce incentives to early withdrawal need to come along with active labour market policies for the less skilled and with less restrictive EPL to incite firms hiring such workers.
3.58 The labour market therefore needs a global reform that should combine easing EPL with reduction in labour cost for the low paid, improvements in the efficiency of the public employment services and suppression of the incentives to withdraw from the labour market. Although the authorities seem to be aware of the need of a global reform, judging by their actions in various fields (relaxation of some aspects of the social modernisation law and of the 35 hours legislation, improvement of job search monitoring), policies seem to continue to concentrate on some aspects of the reform, mainly on ALMP at the moment and less on more fundamental reform that addresses the unintended perverse effects of EPL. Policies also leave aside the problem of inactivity traps caused by an inconsistency of family policies and employment objectives. Finally, policymakers have difficulty in convincing the public that increases in the real level of the minimum wage are more likely to create unemployment than to increase aggregate demand.
ANNEX 3.A1. EVOLUTION OF THE EMPLOYMENT PROTECTION LEGISLATION
Despite some exceptions, the rigidity of EPL has gradually increased since 1970:
Legislation on Legislation on temporary contract permanent contract 1973 Law The burden of proof in the case of a dismissal is shared between the employer and the employee. 1975 Law Introduction of prior administrative authorisation for dismissal for economic reasons. 1982 Ordonance The permanent contract is legally the reference contract. The use of temporary work is restricted. Introduction of a premium imposed on firms when a fixed term contract expires and is not transformed into a permanent one. 1986 Law Reinforcement of the dismissal procedure: increase in the notice period, consultation of enterprise committee (works council), obligation on the employer to have an interview with the employee. Suppression of the prior administrative authorisation for dismissal for economic reasons. 1989 Law Definition of a strict framework for dismissals for economic reasons. An economic dismissal is a dismissal for a reason which is not inherent to the person but is caused by economic difficulties or technological changes. An economic layoff gives right to have access to a reclassification program, with a priority to be re-hired. Reinforcement of the procedure: Obligation for the employer to inform the enterprise committee about employment developments over the next year. 1990 Law Statement that temporary contracts should not be used to fill a permanent job linked to permanent activity of the firm and definition of the case in which they can be used. Strict definition of the framework of the contract (duration, payment). 1991 Law Framework for the employee to have legal advice. 1992 Jurisprudence The reason for an economic dismissal can be a reorganisation of the firm. The firm needs to adapt employees to the evolution of their jobs. 1995 Jurisprudence The notion of an economic dismissal justified by a reorganisation of the firm is abandoned, replaced by the notion of competitiveness safeguard. Jurisprudence Obligation in the case of a dismissal of more than 10 workers to put in place a "plan social", that should include precise measures so as to ease the reclassification of workers and avoid the number of dismissed workers. 1997 Jurisprudence A financial rationalisa- tion of the management does not justify economic dismissal. 1998 Act of It is in the power of the 30 December judge to assess the 1998 seriousness of the reason of the economic layoff. 1999 Jurisprudence The burden is on the employer to prove that he met his obligation to find another job for the worker. 2000 Jurisprudence The employer should put in place all the measures, within the means of the firm, in order to maintain the level of employment and to facilitate internal re-assignment. 2002 Law (see Increase in the Reinforcement of the below) premium. procedure for economic dismissal. Reinforcement of the measures that should be taken by the employer to avoid dismissals. Narrowing of the reasons that can justify economic dismissal (summary of the jurisprudence). 2003 Jurisprudence Ease of the use Suspension of some of "common use articles of the 2002 Law temporary contract". (see below). In case of a dispute about the recourse to "common use temporary contract", the judge should only check that it is common practice not to employ a worker on a permanent contract for that job.
In 2002, the strictness of EPL has been reinforced by the "Loi de modernisation sociale", but part of the provisions has been reformed in 2003 and 2004 and new ones have been introduced by the "Plan de cohesion sociale". Main points introduced since 2002 are the following ones:
Legislation on dismissal (permanent contract) Reassignment actions Before putting in place a collective dismissal (more than 9 workers dismissed), firms should: --have implemented working time reduction; Suppressed. --have started a "plan de sauvegarde de l'emploi", Maintained. which consists of efforts to re-assign the worker to another job of the same category or of a lower category (with the agreement of the worker) in the group to which the firms belongs. The employer also has the obligation to do his best to train and to adapt the worker so that he is able to cope with job changes. Offers made to the worker should be written and precise. A dismissed worker has priority to be hired Maintained. in the firm for a period of 12 months. Firms with more than 1 000 employees have Maintained. to offer a "re-assignment leave" (conge de reclassement) during which the worker can be re-assigned without any interruption in his contract. It lasts between 4 and 10 months during which a unit must be set up in the firm to take charge of the re-assignment of workers. The employee receives an allowance, which is at least 65% of yearly gross wage (and can not be less than 58% of the SMIC). It is paid half by the firm, half by the state (the Fonds national pour l'emploi). Firms with more than 300 employees have to Introduced by the negotiate every three years on the evolution Social Cohesion of employment and skills among the firm Plan (SCP) A "re-assignment leave" is created for any Introduced by the employee in a firm with less than 1 000 SCP employees, managed by the PSE and the houses for employment and financed by the firm, the unemployment insurance system and the State. Re-hiring if asked by the judge when the Introduced by the procedure has not been correctly followed SCP can not be required if the firm has closed or if there is no job availability. Procedure The Law states the number of meetings with Suppressed. the enterprise committee (works council) and the periods between them. The joint production committee has a right to be advised by an expert and to oppose the dismissal plan. In firms with more than 50 salaries, the Maintained. procedure includes regular consultation of the enterprise committee. In case of disagreement in the termination of Suppressed. a firm activity that implies the dismissal of hundred workers, a mediator should intervene and make a proposal to both parties. When a collective dismissal leads to a Maintained. reduction in activity that, in return, impacts a subcontracting firms, those firms should be informed. In the case of a termination that implies the Suppressed. dismissal of more than a hundred workers, the employer must present the social and territorial consequences of the termination to the Labour Inspectorate (a verifier). The procedure for a dismissal for economic Introduced by the reason can be defined by contractual SCP agreements. Some delays of the procedure are shortened. Definition of a dismissal for economic reason When examining the legitimacy of a collective Maintained. dismissal, the administration can take into account financial position of the group to which the firm belongs. Professional quality can not be a criterion Suppressed. to decide which worker is going to be dismissed. Legislation on temporary contracts Premium that is paid by the firm on expiry Maintained, but if of the contract (if the contract is not social partners converted to a permanent one) increased agree and if the from 6% of gross wage to 10%. worker is given access to special training, the premium is 6%. An employee can terminate a temporary Maintained. contract if he has found a permanent one. Firms have to inform employees on a Maintained. temporary contract about permanent jobs offers.
ANNEX 3.A2. COLLECTIVE BARGAINING REFORM--APRIL 2004
Collective bargaining has been reformed in 2004 with two major innovations. First, the principle according to which an agreement can only improve the employees' rights laid down in an agreement of a higher level was relaxed in certain cases, in particular in respect of hours worked. (The hierarchy consists of intersectoral agreements, sector agreements and company level agreements.) Second, it changed the majority principle: it is not sufficient anymore for an agreement to be valid that it has been signed by just one trade union with representative status.
New relationships between norms
The principle of favourability to the employee according to which an agreement can only improve the employees' rights laid down in an agreement of a superior level is modified. A sector-level agreement may deviate from the provisions of an intersectoral agreement unless such derogation is expressly forbidden by that intersectoral agreement. A company-level agreement may, in turn, deviate from all or part of a sector- level agreement, again unless such derogation is expressly forbidden at the higher level. There are also other restrictions on the right for lower-level agreements to depart from higher ones. The favourability principle is retained for minimum wages, job classifications, supplementary social protection measures, and multi-company and cross-sector vocational training funds.
The previous 'hierarchy of norms' in terms of collectively agreed provisions remains in force for agreements reached before the new law came into force.
Validity of agreements and majority principle
The legislation introduced a 'majority principle' for collective agreements to be regarded as valid. Previously, it was sufficient for an agreement to have been signed by at least one trade union with representative status. The new majority principle will be applied differently depending on the level of negotiations:
* At intersectoral-level, an agreement is valid in the absence of opposition by a numerical majority of trade union organisations with representative status.
* At sector level, two variations of the majority support principle are suggested, and the social partners must choose one of them in a sector-level agreement. The first model that of the intersectoral-level. Under the second model, the agreement must be signed by one or more unions representing the majority of the employees in the relevant sector. If there is no agreement stipulating conditions for the validity of sector-level agreements, the opposition of a majority of unions is applicable to the sector.
* At company level, two different variations of the majority principle are available, and the social partners have to choose between them and enshrine this in a sector-level agreement. In the first model, in order to be valid a company-level agreement must be signed by one or more unions that received at least 50% of the votes cast in the first round of the most recent works council or workforce delegate elections. If no union holds this majority, the agreement must be approved by a majority of employees in a vote. In the second variant, an agreement is valid if not opposed by non-signatory unions that received at least 50% of the votes cast in the first round of the most recent works council or workforce delegate elections.
Collective bargaining in SMEs
To facilitate the conclusion of collective agreements in small and medium-sized enterprises (SMEs) with no trade union delegates, the law allows the social partners to reach agreements on methods in each sector that enable firms to conclude agreements with elected staff representatives or with employees designated by a representative union.
Table 3.1. Average duration (1) of temporary contracts by sector (in months) 1996 1997 1998 1999 Total activity 2.7 2.6 2.6 2.6 Industry, including energy 3.6 3.4 3.5 3.7 Construction 4.2 3.9 4.2 4.2 Services 2.4 2.4 2.4 2.3 (1.) Data exclude workers employed on a temporary contract through an agency, where the average placement is of shorter duration. The duration is unknown for 18% of the sample population. Source: DARES, DMMO/EMMO. Table 3.2. The anatomy of a typical work year for full-time dependent employees, 2002 Decomposition of average annual hours actually worked by full-year equivalent into its components Extra hours on main job Usual = Annual Average weekly Overtime Hours on hours of weekly hours of + additional work hours work in variable jobs on all the main hours jobs job (e.g. flexible hours) + (b) others (a) = =(c)+(d) (b)*(f) +(e) (c) (d) (e) Hours Weekly hours worked Austria 1 657 41.9 40.1 1.5 0.4 Belgium 1 600 39.8 39.3 0.3 0.2 Switzerland 1 875 44.1 40.9 2.9 0.3 Germany 1 679 41.3 39.9 1.3 0.2 Denmark 1 597 40.5 39.1 0.8 0.6 Spain 1 712 40.7 40.4 0.1 0.2 Finland 1 589 40.9 39.2 1.4 0.4 France 1 569 38.6 37.7 0.8 0.1 Greece 1 851 41.5 41.0 0.1 0.3 Hungary 1 821 41.5 40.9 0.4 0.2 Ireland 1 755 39.9 39.5 0.2 0.2 Iceland 2 022 50.7 47.3 1.8 1.6 Italy 1 594 38.8 38.5 0.1 0.1 Luxembourg 1 680 40.1 39.5 0.5 0.1 Netherlands 1 604 40.5 38.9 1.4 0.2 Norway 1 517 41.1 38.5 1.8 0.8 Portugal 1 729 41.3 40.3 0.3 0.8 Sweden 1 503 41.8 39.9 1.3 0.6 United Kingdom 1 805 44.2 43.3 0.7 0.2 Annual Holidays Absences weeks and due to non worked vacation holiday weeks reasons (f) = 52 - 1(g) + (h)] (g) (h) Weeks worked/not worked Austria 39.5 7.3 5.2 Belgium 40.3 7.1 4.6 Switzerland 42.6 6.1 3.3 Germany 40.6 7.8 3.6 Denmark 39.4 7.4 5.2 Spain 42.1 7.0 2.9 Finland 38.9 7.1 6.0 France 40.7 7.0 4.3 Greece 44.6 6.7 0.7 Hungary 43.9 6.3 1.8 Ireland 43.9 5.7 2.3 Iceland 39.9 6.2 5.9 Italy 41.1 7.9 2.9 Luxembourg 41.9 7.5 2.6 Netherlands 39.6 7.6 4.8 Norway 37.0 6.5 8.5 Portugal 41.9 7.3 2.8 Sweden 36.0 6.9 9.2 United Kingdom 40.8 6.6 4.6 Source: OECD (2004). Table 3.3. Evolution of working time legislation Before the 35 hour Aubry laws week. (1998-2002). Standard working 39 hours per week. 35 hours per week. time. Level of overtime 130 hours per 130 hours per worker before a rest period worker per year. per year, but more is compulsory. during the transition period. Overtime premium. 25% between 40 25% between 36 and and 43 hours and 43 hours (41) and 50% 50% above beyond 43. 43 hours. For firms with less 10% between 36 and than 20 employees. 43 hours. "leave" account Did not exist. Can be used for leave or training only. (i.e., to store accrued but unused leave). Fillon law (2003). Act of 21 March 2005. Standard working 35 hours per week. 35 hours per week. time. Level of overtime Fixed by branch level 220 hours except in before a rest period agreement. case of branch or is compulsory. firm level agreement. Otherwise, 180 hours. Overtime premium. Premium can be Fillon's framework negotiated at the extended until 2008. branch level between 10 and 25%, 25% For firms with less otherwise. than 20 employees. 10% between 36 and 43 hours. "leave" account Part of unused leave A collective branch can be paid (5 days level agreement can (i.e., to store per year max) or decide that the leave accrued but unused stored (22 days account can be paid leave). per year max.) and (with no limit on the used within five years. number of days); otherwise, used for leave or training. Repeal of the annual ceiling of 22 days. Table 3.4 Decomposition of real labour cost growth between real wage and social security contributions Average annual growth Contribution of social Contribution of real contributions growth wage growth (%) (%) 1970-1985 2.1 1.0 1986-2003 0.8 0.3 1986-1992 1.2 0.5 1993-1997 0.4 0.1 1998-2003 0.8 0.2 [1986-2003]-[1970-1985] -1.3 -0.7 Share of real Total real labour wage growth in cost growth (%) total growth (%) 1970-1985 3.1 68 1986-2003 1.1 76 1986-1992 1.6 72 1993-1997 0.5 80 1998-2003 1.0 83 [1986-2003]-[1970-1985] -2.0 63 Source: OECD.
(36.) Defined as the unemployment rate at which inflation is stable.
(37.) The differences in hours worked per capita (across countries and over time) can be decomposed into: (1) differences in hours per employee; (2) differences in the employment rate (the ratio of employment to working-age population); and (3)differences in demography (the ratio of working-population to total population). Hours worked per employee depend on the incidence of part-time work and on working time for full-time workers.
(38.) Except for the suppression of the prior administrative authorisation for dismissal for economic reason in 1986.
(39.) The "35 hours legislation" is generally considered to be the set of the three laws, the loi de Robien, the first and the second loi Aubry.
(40.) M. Durand and J. Martin, The 35-hour week, Portrait of a French exception, in: OECD Observer no. 244, September 2004.
(41.) During the transition scheme, premium between 35 and 39 hours was limited to 10%.
(42.) In 2000, monthly earnings were kept constant for 98% of those workers who passed to the 35 hour week. (Pham, 2002).
(43.) CFDT, confederation francaise du travail; CFTC, confederation francaise des travailleurs chretiens; CFE-CGC, confederation francaise de l'encadrement--confederation generale des cadres; CGT, confederation generale du travail ; CGT-FO confederation generale du travail--force ouvriere.
(44.) Before that, only a suppression of the benefit was possible. As a result, the sanction was rarely imposed.
(45.) The firm is exempt from the contribution if the worker in question was over 45 when hired and was hired after May 2003.
Anne, D. and Y. L'Horty (2002), "Droits connexes : transferts sociaux locaux et retour a l'emploi", Document EPEE No 02-04.
Artus, P. and G. Cette (2004), "Productivite et Croissance", Rapport du Conseil d'Analyse Economique No. 48, La Documentation Francaise.
Balmary, D., C. Chevrier-Fatome, et B. Simonin, (2004), Rapport de l'Instance d'Evaluation de la Politique de l'Emploi et Recours a des Operateurs Extemes, La Documentation Francaise.
Behaghel L., B. Crepon and B. Sedillot (2004) : "Contribution Delalande et transitions sur le marche du travail", Economie et Statistiques, No. 372, pp 61-88.
Berger, E., M. Defosseux, K. Even, A. Fleuret, S. Mas and R. Sanchez (2004), "Les beneficiaires des politiques d'emploi, trois ans apres la sortie d'un contrat aide: entre consolidation professionnelle et reinstallation dans le chomage", Premieres informations et premieres syntheses, No. 32.1, DARES, Paris.
Bertola, G., T. Boeri and S. Cazes (2000), "Employment Protection in Industrialized Countries: the Case for New Indicators", International Labour Review, Vol 139 N[degrees] 1.
Bilan de la negociation collective en 2003 (2004), Ministere de l'Emploi, du Travail et de la Cohesion sociale, Editions Legislatives, Paris.
Biscourp P., N. Fourcade (2003), "Peut-on mettre en evidence l'existence de rigidites et la baisse des salaires a partir de donnees individuelles? Le cas de la France a la fin des annees 90", INSEE working paper G 2003/09.
Blanchard, O. and A. Landier (2002), "The Perverse Effect of Partial Labour Market Reform: Fixed-Term Contracts in France", Economic Journal, Vol. 112, No. 480, pp. F214-244.
Blanchard, O. and J. Tirole (2003) "Protection de l'Emploi et Procedures de Licenciement" Rapport du CAE No. 44, La Documentation Francaise, Paris.
Blanchard, O. and J. Wolfers (2000), "The Role of Shocks and Institutions in the Rise of European Unemployment: the Aggregate Evidence", Economic Journal, Vol. 110, No. 462, pp. C1-33.
Bommier, A., T. Magnac and M. Roger (2003), "Le marche du travail a l'approche de la retraite entre 1982 et 1999, evolutions et evaluations", Revue Francaise d'Economie, No 18, p. 23-82.
Bonnechere, M. (2002), "Les transformations de l'ordre public salarial", Communication au colloque EPEE-DARES, "Emplois flexibles, salaires rigides?".
Brandt, N., J. M. Bumiaux and R. Duval (2005), "Assessing the OECD Jobs Strategy: Past Developments and Reforms", OECD Economics Department Working Paper No. 429.
Bunel, M., T. Coutrot and S. Ziberman (2002), << Le Passage a 35 Heures vu par les Employeurs >>, Premieres informations et premieres syntheses, No. 17.2, DARES, Paris.
Cahuc, P. (2003), "Pour une Meilleure Protection de l'Emploi", Document de travail du COE, No. 63.
Cahuc, P. and F. Kramarz (2004), "De la Precarite a la Mobilite: Vers une Securite Sociale Professionnelle", Rapport au Ministre d'Etat, Ministre de l'Economie, des Finances et de l'Industrie et au Ministre de l'Emploi, du Travail et de la Cohesion sociale.
Cahuc, P. and F. Postel-Vinay (2002), "Temporary Jobs, Employment Protection and Labor Market Performance" Labour Economics, (9) 1 (2002) pp. 63-91.
CERC (2005), << La securite de l'emploi face aux defis des transformations economiques economiques >>, La Documentation Francaise, Paris.
Cour des comptes (2003), Chapitre I, Rapport au President de la Republique.
Crepon, B., M. Dejemeppe and M. Gurgand (2004), "Counseling the unemployed: does it lower unemployment duration and recurrence?", mimeo.
DARES (2003), "Les Politiques de l'Emploi et du Marche du Travail", Collection Reperes, La Decouverte, Paris.
De Virville, M. (2004), "Pour un Code du Travail plus Efficace", Rapport au Ministre des Affaires sociales, du Travail et de la Solidarite.
Desplatz, R., S. Jamet, V. Passeron and F. Romans (2003), "La moderation salariale en France depuis le debut des annees 1980", Economie et Statistique no367.
European Commision (2003), "Employment in Europe 2003, Recent Trends and Prospects".
Fougere, D. (2000), "Accompagnement des Chomeurs et Sanctions: leurs Effets sur le Retour a l'Emploi", in plein Emploi, Rapport du Conseil d'Analyse Economique No. 30, La Documentation Francaise.
Goux, D. (1997) "les salaires nominaux sont-ils rigides a la baisse?" INSEE Working paper.
Jugnot, S. (2002), "Combien d'emplois crees par la reduction du temps de travail?" in Donnees Sociales, La Societe Francaise, p. 255-262.
Kramarz, F. and M-L. Michaud (2004), "The Shape of Hiring and Separation Costs", IZA Working Paper, 1170.
Laroque and Salanie (2000), "Une Decomposition du Non-Emploi en France", Economie et Statistique No. 331, p. 47-66.
Le SMIC, salaire de croissance (1999) rapport du CSERC, la Documentation Francaise.
Marimbert, J. (2004), "Le Rapprochement des Services de l'Emploi" Rapport au Ministre des Affaires sociales, du Travail et de la Solidarite.
Martin, J., M. Durand and A. Saint-Martin (2003), "La Reduction du Temps de Travail : une Comparaison de la Politique des "35 Heures avec les politiques d'autres Pays Membres de l'OCDE", Audition OCDE Assemblde Nationale du 22 Janvier 2003.
OECD (2002), Employment Outlook, Paris.
OECD (2003), Economic Survey France
OECD (2004) "Assessing the OECD Jobs Strategy: Past developments and Reforms"
OECD (2004a), Employment Outlook, Paris.
OECD (2004b), Benefits and Wages, Paris.
OECD (2005), Ageing and Employment policies--France, Paris.
Olievera, A. And V. Ulrich (2002), "L'incidence des 35 heures sur le temps de travail", Premieres informations et premieres syntheses, No. 07.1, DARES, Paris.
Piketty, T. (1998), "L'Impact des Incitations Financieres au Travail sur les Comportements Individuels: une Estimation pour le cas Francais", Economie et Prevision No. 132-133, p. 1-35.
Pisani-Ferry, J. (2003), "The surprising French employment performance: What Lessons?" CESifo Working Paper No. 1078.
Pham, H. (2002), "Les modalites de passage a 35 heures en 2000", Premieres informations et premieres syntheses, No. 06.3, DARES, Paris.
|Printer friendly Cite/link Email Feedback|
|Publication:||OECD Economic Surveys - France|
|Date:||Jun 1, 2005|
|Previous Article:||Chapter 2: making the fiscal system sustainable and more efficient.|
|Next Article:||Chapter 4: product market competition and economic performance.|