From social assistance to self-sufficiency; low paid work as stepping stone

Introduction
Welfare reforms, both in the US and in European countries, focus on stimulating employment among welfare recipients. The underlying premise of labour market activation policies is that benefit recipients will be better off in the long run if they take a job (Brown 1997). Jobs that do not pay a living wage are seen as an effective way to enhance employability and as stepping stones towards jobs that provide higher wages or more hours and thereby self-sufficiency (Johnson and Corcoran 2003; Pavetti and Acs 2001). In addition, long term welfare receipt is seen as a trap that will further worsen the situation of recipients (Mood 2013). However, little is known about whether a part-time or low paid job is an effective stepping stone from social assistance towards (long-term) self-sufficiency. Some researchers find that if former welfare recipients find a (low paid) job, their income increases with tenure and work experience (Loeb and Corcoran 2001). Others argue that most welfare recipients will obtain poorly paid jobs with few training and promotion opportunities (Andress and Lohmann 2008; Burtless 1995) and therefore their income from work will not lift them above the poverty line (Connolly and Gottschalk 2000; Edin and Lein 1997; Newman 1999).
In this paper we focus on social assistance recipients in the Netherlands and the role of (low paid) work as a stepping stone from social assistance towards self-sufficiency. In earlier research, mixed results were found on whether or not low paid work can function as a stepping stone towards self-sufficiency. Possibly, these mixed results can be explained by differences between individuals or between (low paid) jobs. Therefore, in this paper, we study for whom low paid work is (especially) effective and which (low paid) jobs are the most effective stepping stones towards self-sufficiency.

Data and methods
For this paper we use administrative data on the entire population of the Netherlands. We select all people who received (any income from) social assistance in January 2010 and follow these people until 2015. About 9% of social assistance recipients in the Netherlands are combining welfare and work; they work a part-time job but earn less than the social minimum, therefore they also remain dependent on social assistance.
We estimate discrete time event history models in which we estimate whether people exit from social assistance to self-sufficiency. Our main independent variable is whether people have a (low paid) job. These people receive an income from work in a month, but their income is below the social minimum, therefore they also remain dependent on social assistance. We include personal characteristics (gender, age, ethnicity, educational level, household situation) and individual labour market histories (work experience and duration of social assistance receipt). Of the people who have a (low paid) job in a month, we have several job characteristics. We know whether they are interns, work in sheltered employment are on-call workers, temporary agency workers or have a regular employment contract. In addition, we have characteristics (number of employees, turnover rate and share of employees with non-regular employment contract) of the organisation where they work.
We include labour market activity in low paid work in our models as a time-varying covariate. Therefore we do not only compare between working and inactive people but also between working person-months and person-months in which the same individual was inactive. We estimate the effect of low paid work in a certain month on the odds of experiencing an event. For 18% of the individuals in our dataset, there are both inactive and working person-months included, that is, there is variation in activity within individuals. As a sensitivity analysis we re-estimate the models on only these individuals. That is, we estimate the effect of low paid work in a month on self-sufficiency, within a group of individuals who are all at some point active in low paid work.
We thus estimate whether low paid work has an effect on exits from social assistance towards self-sufficiency. Thereby we are interested in the moderating role of personal characteristics; that is, for whom is a low paid job (especially) effective? In addition, we are interested in the role of job and organisation characteristics; which low paid jobs function as stepping stones towards better paid work? To answer these questions we estimate a number of discrete time event history models, in which we estimate the odds of exit from social assistance towards self-sufficiency from labour market activity, personal characteristics, labour market history, job characteristics and interactions.

Results and conclusions
In this paper we study whether this low paid work increases the opportunities of social assistance recipients to become completely independent of social assistance. In discrete time event history models we estimate the effect of low paid work on subsequent sustainable exits from social assistance towards self-sufficiency. We find that low paid work increases the odds of experiencing an event; that is, of exiting social assistance, remaining independent of social assistance for at least six months and earning an income on the labour market (rather than moving to another benefit type or moving in with a working partner). This finding is in line with our expectations based on human capital theory, social capital theory and signalling theory. If social assistance recipients are active in low paid work, this will increase their work experience and skills, their access to valuable labour market information and they will signal their reliability and willingness to work to (future) employers. Therefore, they will be more likely to find work with an income above the social minimum and thus become independent of social assistance.
Secondly, we focus on individual differences in the effect of low paid work. We hypothesised that especially people with little work experience and long term social assistance receipt could gain relevant (up to date) human capital and social capital from low paid work. Also signalling their reliability, productivity and willingness to work will be especially important for these groups. Our results confirm these hypotheses, we find that low paid work especially increases exits from social assistance to self-sufficiency for people with little work experience and people with long term social assistance receipt.
Low paid work is expected to be less effective for higher educated people as they can gain less relevant human capital in low paid work and have more other (and better) means (diplomas) to signal their productivity. This is confirmed by our results; especially for lower educated people low paid work functions as a stepping stone towards self-sufficiency.
Ethnic minorities have less access to social capital and labour market information in their family and friendship networks. As low paid work will provide social assistance recipients with labour market information and connections, this might be especially effective for ethnic minorities. However, we find no evidence for this, the stepping stone effect of low paid work is smaller for ethnic minorities than for the native majority. This finding is in line with the expectations based on the dual labour market theory; ethnic minorities in low paid work are more likely to work in the secondary labour market or the enclave labour market which constrains their opportunities for upwards mobility towards better paying jobs and self-sufficiency.
Thirdly, we focus on job and organisation characteristics to investigate which types of low paid work are the most effective stepping stones towards self-sufficiency. We hypothesised that there are more opportunities for upward mobility on the internal labour market in larger organisations. We find some evidence for this; compared to a low paid job in an organisation with less than 25 employees, low paid work in larger organisations has a larger effect on exits from social assistance towards self-sufficiency. This is very similar to the findings of Grün and colleagues (2011) for Germany and Austria. We also hypothesised that organisations with high turnover rates and high share of people with non-regular employment contracts are more likely to be secondary labour market organisations, which offer less opportunities for training, promotion and thus upward mobility. We find, however, no evidence for this. The effectiveness of low paid work as a stepping stone does not depend on the turnover rate in an organisation, while working for organisations with non-regular employment contracts is more effective than working in organisations with no non-regular employment.
When we focus on differences between job types, we find that especially temporary agency work functions as a stepping stone towards self-sufficiency. Social assistance recipients with a temporary agency job are more likely to become self-sufficient than those in sheltered employment, internships, on call work or regular employment contracts. We did not expect this, but Grün and colleagues (2011) find, also to their surprise, exactly the same effect for both Germany and Austria. They study unemployed men who start a low wage job, and find that those who do temporary agency work have the highest risk of upward mobility towards higher paid work, but also the highest risk of returning into unemployment. These outcomes might be explained by the fact that employers use low paid work, and especially temporary agency work, as a way to screen newly hired workers. After a short screening period organisations can easily get rid of temporary agency workers. However, if workers prove themselves, they can get an employment contract directly from the employing organisation, which will generally lead to an increase in their wage, or they will get a new position via the temporary employment agency with a higher wage or more hours.

In conclusion, we find that a period of working but earning less than the social minimum is an effective stepping stone towards sustainable self-sufficiency. Especially low educated people, people with little work experience and long term social assistance recipients increase their opportunities to find work with an income above the social minimum if they are active in low paid work.