Attrition, induced quits and layoffs

Malmö Cigar Factory was one of the Tobacco Monopoly’s three establishments for cigar and cigar-cigarette production. The factory employed about a third of the workforce in the branch and a fifth of the company’s whole workforce at the end of 1920.¹ The development of employment at Malmö Cigar Factory is fairly similar to that of the company as a whole, as evident in figure 8.1. First a rapid expansion, with the peak reached at the end of 1919 when the factory employed more than 1,200 workers, and thereafter a slight contraction, followed by a major reduction of the workforce. The decrease continued at a slower pace until 1927 when a more dramatic downturn occurred, followed in 1929 by a temporary expansion. By then, the number of workers had been reduced by over 60 percent since 1920.

Figure 8.1 also relates the stock of workers at Malmö Cigar Factory at the end of each year to the flow of workers – hirings and exits. Exits are here defined as all those workers who left the company, voluntarily or involuntarily.

This figure shows a marked reduction in the inflow of workers as the depression

1 Aktiebolaget Svenska Tobaksmonopolets verksamhet år 1920: Styrelsens förvaltningsberättelse, p 24.

.

Figure 8.1 The workforce at Malmö Cigar Factory – stocks and flows, 1915-1939

-600 -400 -200 0 200 400 600 800 1000 1200 1400

1915 1918 1921 1924 1927 1930 1933 1936 1939

Total number of workers Hirings Exits

Note: Total workforce is measured at the end of each year. The number of exits in this figure is inverted to represent the outflow of workers.

Source: MS, FHK, Matriklar över slutade arbetare, D4A: 1-8.

began to be felt in 1920. Thereafter the inflow of workers was kept at a low level. The outflow of workers was low throughout the whole period with the exception of two years: 1921 and 1927.

The personnel records from Malmö Cigar Factory make it possible not only to reconstruct flows into and out of the factory over time, but also to look at why workers left the factory. The exit reasons from 1920 to 1928 are summarized in table 8.1, which shows that layoffs ‘due to shortage of work’ accounted for more than a third of the total number of exits and that quits ‘at own request’ were about as important.² Pensions, as expected, had only a limited significance over the period as a whole. However, it should be noted that the senior workers who were temporarily laid off in October 1921 were not given a specific exit reason in the personnel records, not even when the pension scheme came into effect in 1922, and are found in the category ‘No cause stated with compensation’ in table 8.1. If these exits are taken into account, pensions corresponded to 10 percent of all separations.

Table 8.1 summarizes the gross outflow of workers from the factory by exit reason. However, there was also a certain inflow which has to be taken into account when assessing the importance of various measures. During the period

2 For a more detailed description of the exit reasons used in the personnel records of Malmö Cigar Factory, see section A2.7.

in question (1920-1928), 142 workers were hired at the factory. As mentioned in chapter 2, an employer has basically three options when reducing the number of workers: attrition, buyouts and layoffs. Attrition is here defined as all separations, except for buyouts and layoffs, minus the number of new hirings.

Table 8.1 Exit reasons at Malmö Cigar Factory, 1920-1928

Reason Number of

exits

Share of all exits (in percent)

Classification

Shortage of work 364 36 L

Discipline offence or incompetence 9 1 A

Dismissal – unspecified cause 12 1 A

At own request

without compensation 257 25 A

with compensation 105 10 B

Pension

old age 49 5 A

early retirement 6 1 B

Disability & sickness 12 1 A

Death 32 3 A

Transfer to other factory 12 1 A

Other 3 0 A

No cause stated

without compensation 93 9 A

with compensation 67 7 L

Sum 1,020 100

Note: The category ‘Discipline offence or incompetence’ refers to dismissals due to lack of discipline, drunkenness, absence, theft and the like. The classifications in the fourth column represent: B = Buyout, L = Layoff, A = all other separations (gross attrition).

Source: MS, FHK, Matriklar över slutade arbetare, D4A: 1-8.

Buyouts are defined as separations where the employer induces the worker to quit by offering an amount of money, whereas layoffs are cases where employment contracts are terminated due to shortage of work.

Column four in table 8.1 shows how various separation causes have been classified. The classification is not always straightforward. Retirements due to old age are, for example, commonly viewed as a form of attrition. However, in the present case it should be remembered that there was, until the late autumn 1921, neither a pension scheme nor a certain age when workers were expected to resign.³ The pension scheme was introduced to alleviate the consequences of job losses for senior workers at the company. Those who were temporarily laid off in October 1921 awaiting the pension scheme to come into force have therefore been counted as laid off. Another tricky aspect is how to classify workers who quit ‘at own request’ and, as shown in table 8.1, sometimes were given compensation and sometimes not. Those who were compensated are considered as induced quits in this study. It may, however, be discussed how ‘voluntary’

these quits were, since layoffs were likely to follow if too few workers volunteered.⁴ An interesting indication of the sometimes blurry distinction between induced quits and layoffs was found in 1924, when the company offered a number of workers compensation for quitting. These exits were registered as ‘due to shortage of work/at own demand’ in the personnel records.

When classifying the different exits according to the above definitions, and accounting for the inflow of workers, it may be established that attrition represented 38 percent of the workforce reduction, whereas induced quits represented 13 percent and layoffs the remaining 49 percent. Although both attrition and layoffs were important measures, these outflows were distributed unevenly over time.

The bulk of separations classified as attrition were voluntarily quits (‘at own request’). Figure 8.2 shows how the quit rate for male and female workers developed over time at Malmö Cigar Factory.⁵ As displayed in the figure, the quit rate approached 10 percent in 1920 for both men and women, but, as the downsizing process began, the voluntary turnover decreased and fluctuated

3 Although there was a general pension system in Sweden at the time, the compensation amounts in this system were hardly high enough to allow complete withdrawal from the labour force.

Consequently, 47 percent of men over 70 were gainfully employed in 1920. Olsson 1986, p 13.

4 This was, as mentioned, the case in the spring of 1927. The issue of ‘voluntary’ quits has been discussed in general terms by Turnbull 1988, p 207.

5 The quit rate is defined as the number of exits ‘at own request’ (without compensation) divided by the average of the stock of workers at the beginning and end of each year.

around a considerably lower level until the end of the inter-war period.⁶ This pattern can be attributed to external as well as internal factors. Looking at the external labour market, there is some evidence indicating that personnel turnover decreased in the inter-war period due to the high unemployment levels.⁷ It may also be established that tobacco workers had particularly few alternative employment opportunities and fairly decent wages. However, even if a tobacco worker got a job offer, he or she had incentives to hesitate since the Tobacco Monopoly compensated redundant workers. Taking another job would imply a lost opportunity to get severance pay. Returning to figure 8.2, it can be noted that for most of the years female tobacco workers seem to have been more inclined to quit ‘at own demand’ than male workers.⁸ This difference, which was particularly pronounced between 1921 and 1926, may indicate the terms on which women participated in gainful employment at the time. If duties towards the family required it, female workers had to quit, irrespective of whether they got compensation for quitting at that moment or not. Another explanation for the gender difference in the quit rate is that the general situation for women in the labour market was somewhat better, since they were typically employed in activities that were less exposed to business cycles.

Although voluntary quits were of decreasing importance they contributed to a more or less continuous outflow of workers from the factory. Induced quits and layoffs were, as shown in figures 8.3 and 8.4, concentrated to certain years.

Induced quits began to be applied on some scale in 1923, when about 20 male workers at Malmö Cigar Factory accepted severance payment. In the following year, a group of female workers of about the same size was induced to leave the factory. The greatest number of induced quits in a single year was recorded in 1927, when this measure was directed to female workers to make room for the male cigar workers who were recalled after the negotiations with the union.

The layoffs were even more concentrated in time than the buyouts. Close to 90 percent of the recorded exits ‘due to shortage of work’ occurred during two years: 1921 and 1927. The existence of fixed costs associated with layoffs is a

6 As a comparison it may be mentioned that quit rates among blue-collars in the Swedish engineering industry exceeded 12 percent most of the years in the 1920s. Before the post-war depression, employers in this industry could experience quit rates above 50 percent. Holmlund 1984, p 25.

7 Holmlund 1984, p 25.

8 Laura Owen reports evidence pointing in the same direction with regard to manufacturing firms in the United States in the 1920s. Owen 2001, pp 53-54.

Figure 8.2 Quit rate at Malmö Cigar Factory, 1915-1939

0 2 4 6 8 10

1915 1918 1921 1924 1927 1930 1933 1936

Men Women

Note: Quit rate is here defined as the number of quits ‘at own request’ during each year per hundred workers. The number of workers has been calculated as the average of the stock at the beginning and end of each year.

Source: MS, FHK, Matriklar över slutade arbetare, D4A: 1-8.

Figure 8.3 Number of induced quits at Malmö Cigar Factory, 1915-1939

0 10 20 30 40 50 60

1915 1918 1921 1924 1927 1930 1933 1936 1939

Men Women

Source: MS, FHK, Matriklar över slutade arbetare, D4A: 1-8.

plausible explanation for the observed pattern. Establishing an order of selection, deciding upon compensation terms, negotiating with the union (or handling protests afterwards) and other matters may not have been much more expensive if the number of layoffs were 10 or 100. Employers therefore prefer layoffs of many workers at a time rather than having an even stream of layoffs.

It is also reasonable that the latter policy have more negative effects on worker morale than mass-layoffs.

Figure 8.4 Number of layoffs due to shortage of work at Malmö Cigar Factory, 1915-1939

0 50 100 150 200 250 300

1915 1918 1921 1924 1927 1930 1933 1936 1939

Men Women

Source: MS, FHK, Matriklar över slutade arbetare, D4A: 1-8.

Table 8.2 Layoff rates for men and women at Malmö Cigar Factory, 1916-1934

Year Men Women

1916 0.0 0.0 1917 0.0 0.0 1918 0.0 0.0 1919 0.0 0.0 1920 2.0 0.2 1921 14.2 26.0 1922 0.0 0.5 1923 0.0 0.3 1924 0.0 3.3 1925 0.0 0.0 1926 0.0 0.0 1927 12.7 8.4 1928 0.0 0.9 1929 2.6 0.3 1930 0.0 0.8 1931 1.3 0.5 1932 0.0 0.0 1933 0.0 0.0 1934 0.0 0.0

Average 1.7 2.2

Note: The layoff rates have been calculated by dividing the numbers of laid-off men and women during each year with the numbers employed at the beginning of each year.

Source: MS, FHK, Matriklar över slutade arbetare, D4A: 1-8.

While figure 8.4 shows that mostly female workers were affected by layoffs, the information in the figure alone does not allow conclusions to be drawn about whether women on average faced a greater risk of being laid off. In order to give a rough idea of the likelihood of male and female workers to be laid off, the absolute numbers of affected workers have been divided by the number of men and women employed at the beginning of each year. The quotas are reported in table 8.2. The revealed pattern is inconsistent; in 1921 women faced a much higher risk, whereas men were more exposed in 1927. On average, however, the layoff rates were higher for women than for men (2.2 percent for women to be compared with 1.7 percent for men).