11 Is a low wage job better than no job?
11.1 The impact of no job on the prospects of getting a job
There is a widespread belief, which underlies much of the design of social welfare policy in English-speaking countries, that any job is better than no job in terms of future prospects for employment and for psychological well-being. For example, the US Department of Health and Human Services explains that “Work First programs share a common philosophy regarding work: any job is viewed as a good job and program efforts should be geared toward helping recipients enter the paid labor force as quickly as possible.” (Holcomb, LaDonna Pavetti and Ricdinger 1998:4, quoted by Gottschalk, 2001:6). The reality is more complex.
A distinction is often made between unemployment and being not in the labour force (inactivity). While the distinction is clear in principle, in practice the line between the two states is blurred. Dropping out of the labour force is one response to unemployment, especially for young people and for mothers. In 1997, there were as many young males who were neither in the labour force nor in education as there were unemployed, in the Netherlands, Sweden and the US (Ryan, 2001).
Non-employment can reduce the probability of getting a job in the next period, or reduce the wage of any job offered, if employers believe that it is signaling lack of motivation or lack of skills. It might signal the latter if those in the un/non-employment pool are the ones that have been rejected by other employers. Unemployment may also have a scarring effect, whereby a person’s work productivity is actually damaged by the experience of unemployment (through psychological damage and the depreciation of work skills). Empirically, it is very difficult to distinguish whether unemployment causes poor labour market prospects, or whether people become and stay unemployed because they have unobserved low-productivity characteristics. What is clear is that having been unemployed reduces the prospects of a young person obtaining subsequent steady work, and most probably reduces pay when employment is found. Job loss for the least educated leads to lower probabilities of re-employment, higher chances of part-time work and lower earnings (Farber, 1999). There is more tentative evidence that unemployment does damage that can last a number of years in terms of lower wages and greater employment instability.
In a careful theoretical and empirical analysis, Stewart (2002) looks directly at the comparison between being unemployed and holding a low wage job, on both wages and employment one and two years later. He uses UK data from the BHPS, and includes people aged 18-65. An interesting dimension of this paper is that Stewart is able to identify the impact of each on unemployment and a low wage job, while holding constant a range of personal attributes. Two of his conclusions are especially pertinent here. The first is that the impact of being unemployed in period 1 on the probability of being unemployed in period 2 is not statistically different from the impact of being in a low wage job. In coming to this conclusion, he excludes from the analysis people who were continuously unemployed over the whole period. The second is that being in a low pay job in period 1 has a similar impact as being unemployed on the probability of being in a low pay job in period 2. Low paid work and unemployment have almost an equally large (negative) impact on the probability of moving to a higher paid job, compared to higher paid employment. Furthermore, being in a low paid job significantly increases the prospect of being unemployed in the next period, compared with being in a higher paying job. “Low paid jobs act as the main conduit for repeat unemployment.” (p 19) On Stewart’s evidence, there is little support for the view that any job is better than no job in improving prospects for future employment or escape from low wage jobs. The negative effects of unemployment on future prospects have been well established (we give some examples below). What is new in Stewart’s work is the direct comparison between a spell of unemployment and a spell of low wage employment.
Evidence of the negative effects of unemployment is abundant. However, analyses that carefully distinguish the independent effect of a spell of unemployment (state dependence) from other causes of unemployment are much less common. Knights, Harris and Loundes (2000) is an example for Australia of separate identification of the effects of past unemployment on future employment prospects. They use data from the late 1980s, drawn from the Australian Longitudinal Survey. They conclude that there is genuine state dependence in unemployment. Unlike the work of Stewart, however, they do not compare this with the state dependence that arises from employment in a low wage job.
Less sophisticated studies of the effects of unemployment on future unemployment and on low wages are available for a number of European countries. For France,
. . . the probability that a young worker gains (or holds) regular employment is significantly reduced by prior unemployment. At the same time, no further damage attaches to employment under fixed-term contracts or participation in labor market programs. Similarly, previous unemployment has been found to increase for German youth the probability of being unemployed---. For British youth, it increases both the probability of entering unemployment and reduces occupational upgrading in early working life. For Swedish youth, it reduces pay when employed.---among (British) males who leave school early, unemployment reduces occupational status, seven years later, only if the relevant spell lasted at least three months. Do the scars fade over time? The durability of adverse effects is uncertain.
Ryan, 2001:48
Outflows from inactivity are lower than from unemployment. “In some countries, many young people shuttle between labor market programs, inactivity, and unemployment, accumulating long spells of joblessness but not of unemployment as they go.” (Ryan, 2001:41).
From a regression estimate of the causes of variance in the hourly wages earned by French youth, Balsan, Hanchane and Werquin (1998: 160-161conclude that labour market history has more impact on women’s than men’s wages.
Being unemployed or in a market sector youth programme corresponds to the lowest wages. -- participation in private sector youth programmes leads to higher wages than a spell of unemployment. However, for a given unemployment duration, having several spells rather than one results in a lower wage. Apparently, queuing on the labour market and searching for a permanent position in the primary sector is more efficient from the wage point of view, rather than going through a series of short duration jobs.... School-to-work employment schemes help young people get into wage earning, “But they might subsequently stay quite a long time in marginal jobs which are often characterised by low wages.
Youth schemes seem to permit better access to jobs, but not to higher wages.
