Selection criteria

between job title and layoff risk, as expected by Cornfield; if not, such a relationship will be weaker or non-existing. Furthermore, groups that have a weak position in the labour market, and which are often crowded in unskilled occupations, are more vulnerable if layoff units are widely defined.¹⁰ This is because job-bumping usually takes place from skilled to unskilled positions.

invested in on-the-job training they will be reluctant to release senior workers.

This is particularly relevant for those workers who have been trained in idiosyncratic tasks, since they are hard to replace if there is a need for more labour in the future.

The intuitive rationale for seniority-based layoffs stemming from basic human capital theory has been criticized by Lorne Carmichael who argues that when general skills grow with experience employers should follow the principle

‘first in, first out’.¹⁵ Firms should use recessions to train young workers to become productive when product demand resumes. If the skills acquired through on-the-job-training are of firm-specific kind, Carmichael’s conclusion is weaker.¹⁶ Then the optimal policy depends on workers’ inclination to quit voluntarily. If personnel turnover is high, it is best for the employer to retain senior workers in downturns. Edward Lazear is another critic of the use of the traditional way of explaining layoffs.¹⁷ He emphasizes that the employer does not only want to keep the most profitable individuals but the most profitable mix of workers, and that this mix depends on technological change and skill requirements. Like Carmichael, Lazear regards reductions at both ends of the age distribution as the typical employer response to shortage of work. But where technological change is rapid and the skills needed are of a general kind, employers will be more interested in retaining young workers. As mentioned, Lazear also sees complementarities between young and old workers. Young workers, with a fresh school education, posses more general human capital and are more updated on the latest technologies, whereas older workers have acquired more firms-specific skills. The existence of such complementarities leads to the conclusion that a mixture of young and old workers is optimal for the employer.

The mixture of young and old workers is also discussed by Lester Thurow, introduced in chapter 2, but with a greater stress on the role of skill transmission workers the employer has to promise not to fire them when they approach the stage when their wages are high. In order to fulfil the contract the employer therefore has to lay off junior workers before seniors in situations with shortage of work. However, it is tempting for employers to cheat – to employ workers while they are young and cheap and to release them when they get old and expensive. This temptation is balanced by the costs associated with having a bad reputation, which in turn depends on a number of factors such as company size and the character of the labour market. Lazear 1981; Ritter

& Taylor 1998.

15 Carmichael 1983, p 1127.

16 Carmichael 1983, p 1130.

17 Lazear 1998, pp 170-173.

for employers’ layoff policies. This theme has been elaborated and formalized by Patricia Reagan, who reasons that experienced workers have a dual function in the firm; they are not only producing output, but also involved in training younger colleagues.¹⁸ These functions are often difficult to separate. Senior workers will refuse to participate in, or even sabotage, the training process when facing the risk of being replaced by newcomers. Employment protection – that is, a layoff policy based on seniority – in Reagan’s model is a way for the firm to induce senior workers to cooperate.

Although there are some theoretical reasons why employers may consider seniority when implementing layoffs, it is often observed that employers want to use seniority as a complementary criterion and ability as the main criterion.

Here, seniority is only used to distinguish between two workers who are equally able. This principle was, for example, codified in the Saltsjöbaden agreement of 1938. It is not hard to understand why employers prefer high-performing individuals to low-performing ones. However, according to Lazear, employers’

eagerness to select workers according to performance is related to the wage system.¹⁹ Remuneration based on piece rates implies that all workers are equally profitable and that employers are indifferent to the order of selection.

Remuneration on an hourly basis implies that the relationship between performance and profits is stronger and that employers will argue for the primacy of ability. Lazear’s reasoning is somewhat simplified, since it does not take into account fixed labour costs and physical restraints. It can be argued that high performing workers are also retained in contexts with piece rates, since this makes it possible to have a lower number of workers employed while producing the same amount of output. Thus, savings can be made on facilities and recruitment costs. Still, this does not exclude the expectation that employers are more eager to keep able workers where wages are time time-based than where they are performance-based.

It is a well established empirical fact that unions generally want layoffs to be governed by straight seniority – according to the principle ‘last in, first out’.²⁰ This was also the case in inter-war Sweden.²¹ There are several explanations for this preference. First, seniority was probably an established fairness norm at the time. Second, if applied strictly, seniority reduced the room for arbitrariness and

18 Reagan 1992.

19 Lazear 1998, pp 189-189.

20 Lee 2004, p 75.

21 Arbetslöshetsutredningens betänkande I. Arbetslöshetens omfattning, karaktär och orsaker. Avgivet av 1926 års arbetslöshetssakkunniga 1931, p 388; Bengtsson 2006, p 111.

favouritism. For individual workers the seniority principle implied increased employment protection, or at least increased possibilities of calculating the likelihood of losing the job in the event of a personnel reduction. Seniority rules were also of strategic importance for the union. If there were no rules governing layoffs, employers could take the opportunity to get rid of unionized workers when conducting workforce reductions. In general, collective action was complicated if workers had to win the approval of foremen all the time in order to keep the job.

The median voter theorem has been used to explain why some unions prefer layoffs to hours-reductions.²² This theory is less convincing if used to explain why unions tend to choose the principle ‘last in, first out’ instead of the principle ‘first in and last in, first out’ since the median union member is unlikely to be laid off irrespective of the choice of policy.²³

Another model for explaining why workers often advocate seniority rules has been proposed by Sangeon Lee.²⁴ He argues that all workers want to avoid arbitrariness in layoff decisions and that they gain by coordinating their actions to persuade the employer to make layoffs rule-based. This is because arbitrariness increases competition among workers and puts a downward pressure on wages. The question is however which rule the workers will come up with.

Lee assumes that there are two kinds of workers – juniors and seniors. The juniors are better off with the principle ‘first in, first out’ and the seniors if ‘last in, first out’ is applied. As the game is set up there is no evident solution. Both parties gain if they can agree upon a common position when negotiating with the employer, although one party gains more than the other. Lee’s explanation of this puzzle is that the past matters. Like the game theorist Thomas Schelling, Lee concludes that “[…] the prominence of a certain solution depends most of all on historical and social factors”.²⁵ Senior workers can convince the junior workers by referring to generally applied norms such as ‘first come, first served’

or to common queue ethics.

Although the most popular principle among unions, straight seniority was not the only existing principle for layoffs in the inter-war period. A rather common alternative was what may be termed ‘the need principle’,²⁶ that is, that

22 Medoff 1979, p 393; Garonna et al 1992. See section 7.2.

23 Lee 2004, p 69.

24 Lee 2004.

25 Lee 2004, p 76.

26 Pfeifer 2007, p 26.

jobs should be allocated to those workers who would be most seriously affected if they lost the jobs.²⁷ The consequences of job losses can in turn be related to individual workers’ ability to find other jobs, their economic situation and obligations towards relatives. The ability to find other jobs may partly be related to age, but also to specific physical disabilities. Being in a troublesome economic situation due to an irresponsible lifestyle may have been hard to use as an argument for employment protection, but referring to the needs of family members was certainly an argument often used when discussing the priority at layoffs. By the mid-1930s, maintenance obligation was mentioned as a selection criterion in the collective agreements of five out of 40 LO-unions and was de facto applied by employers of eight other LO-unions.²⁸

In comparison with the seniority principle, the need principle could be problematic to apply. An example that indicates the complexity has been found by Eva Blomberg in a study of the mining company LKAB.²⁹ During the Great Depression this company used no less than 12 different criteria for layoffs, out of which at least seven may be related to the need principle: number of children, family situation, employment status of father and mother, employment status of brothers, economic situation, health status, possession of real estate and company loans.³⁰ It is not difficult to see why some unions stuck to advocating straight seniority although they found other criteria more attractive in principle.³¹

A further complicating factor with regard to the need principle was gender.

The male-breadwinner ideal prescribed that men, in particular married men, should be protected from unemployment. It is unclear to what extent this ideal

27 There were also some other criteria mentioned in collective agreements. Some unions, particularly in trade, construction, transports and food industry, had managed to include stipulations about preferential job rights for their members. Bengtsson 2006, p 116. As pointed out by Bengtsson (2006, pp 121, 300), privileges for members were most important for unions in industries were with low union density. Where unions had managed to attract the majority of the workforce, as in the tobacco industry, such rules were less meaningful. Place of residence was another selection criterion, sometimes observed in collective agreements from the inter-war era, that can be attributed to union demands. Arbetslöshetsutredningens betänkande I. Arbetslöshetens omfattning, karaktär och orsaker.

Avgivet av 1926 års arbetslöshetssakkunniga 1931, pp 376-377.

28 Bengtsson 2006, p 111.

29 Blomberg 1995, p 234.

30 The other criteria were: skill-level, experience of workplace accidents, carefulness, length of service and absence from work. Blomberg 1995, p 234.

31 Bengtsson 2006, p 111.

was applied in practice in connection with personnel reductions. There is not much evidence found in collective agreements on the matter.³² The public inquiry on married women’s wage work did not focus on personnel reductions, but from the conferences it arranged it appears like some unions encouraged married women to quit voluntarily when there was excess labour.³³ Some unionists were personally in favour of the male-breadwinner norm but preferred a strict application of the seniority norm in practice. One reason for this was the fear of infringements of personal integrity if employers were to scrutinize the family situation of each worker.³⁴

Turning to the actual case studied in this thesis, what behaviour can we expect from the Tobacco Monopoly and the Tobacco Workers’ Union? Not much is known about the layoff procedures and criteria in the industry before nationalization, apart from the fact that the Tobacco Workers’ Union rallied round the seniority principle at the congress in 1906 without much debate. There are no indications of conflicts between junior and senior workers of the kind suggested by some theoretical models.

After 1915, the conditions were changed in two important respects. The first big change – nationalization – meant that general skills became firm-specific. If there had been motives for employers in the pre-monopoly era to apply the seniority principle, the reason for the Tobacco Monopoly was even stronger, but only initially. The second big change – mechanization – pointed in the opposite direction. As technological change made old skills useless and senior workers were no longer important for training young workers, the employer had weaker reasons to consider length of service when implementing layoffs.

Another important aspect of the story is that the Tobacco Monopoly’s position as the sole employer in the industry strengthened incentives to treat the workers in a fair way. This was caused directly by political involvement – manifested in the instructions to the state representatives on the company board – and indirectly by the company’s size and dominating market position.

Reasonably, the costs associated with a bad reputation are higher for a monopsony than for an employer in an industry with a lot of competing firms.

32 Bengtsson has found one agreement according to which married women, whose husbands were fully employed should be the first to go at workforce reductions. Bengtsson 2006, pp 110, 112-113.

33 Frangeur 1998, p 301.

34 Betänkande angående gift kvinnas förvärvsarbete m.m. Avgivet av Kvinnoarbetskommittén 1938, p 314; Arbetslöshetsutredningens betänkande I. Arbetslöshetens omfattning, karaktär och orsaker.

Avgivet av 1926 års arbetslöshetssakkunniga 1931.

Few workers would be interested in entering an industry where there is only one notorious employer. So, the Tobacco Monopoly had reasons to treat the workers fairly. But what did fairness actually mean in this context? Did it imply the application of straight seniority or was the need principle more just? And did fairness imply the participation of the union in decision making? The answers to these questions are not obvious and call for empirical investigation.

Finally, it ought to be recalled that the Tobacco Monopoly, from 1918 onwards, applied a wage system that was partly time-based, partly performance-based. This is of potential importance for understanding the company’s policy regarding the selection of workers for layoffs. The wage system indicates that the workers’ performance was measured. The number of cigars and cigar-cigarettes produced by each person was counted and it was thus possible to observe the productivity of different individuals. Since the wage was only partly related to productivity, the company had a clear incentive to keep high performing workers.