There is growing awareness among policymakers that, in order to break the so-called ‘low pay, no pay’ cycle, labour market programmes must do more than just encourage job entry. To help the unemployed achieve long-term self-sufficiency, they must also support them in work. For a long-time, the UK and the US have provided in-work payments to low-paid workers via Working Tax Credit and Earned Income Tax Credit, respectively. These are available on an ongoing basis, and are intended to sharpen work incentives by increasing the rewards to employment.
While ongoing in-work support is important for making work pay, it is often during the transition from welfare into work that individuals face the greatest difficulties, and the risk of employment exit is particularly high in the period immediately following employment entry. This suggests a possible role for temporary support that can help individuals complete the transition successfully and, with time, become established workers.
Experimental evidence on temporary earnings supplements
A number of interventions involving temporary earnings supplements have been robustly evaluated using randomised control trials. The Minnesota Family Investment Program, the New Hope Project, and the Canadian Self-Sufficiency Project (SSP) all found positive effects on employment, earnings, and income (Michalopoulos 2005). However, these effects faded once the earnings supplements ended. Later programmes combined temporary earnings supplements with employment-related services aimed at helping those eligible to find and retain jobs. In Canada, SSP Plus augmented SSP with employment services and found sustained effects (Robins et al. 2008). In Texas, the Employment Retention and Advancement (ERA) programme showed more mixed results, producing long-term earnings increases in earnings in Corpus Christi, but only short-lived increases in Fort Worth (Hendra et al. 2010). Like its Texan equivalent, UK ERA combined temporary earnings supplements with pre- and post-employment services. There was little evidence of sustained effects for lone parents, but for the long-term unemployed it resulted in positive and sustained employment and earnings impacts (Hendra et al. 2011).
Figure 1. The effect of the UK Employment Retention and Advancement programme on financial year earnings
What has happened to unemployment in the UK?
These positive findings for the UK are of particular interest given the increase in the numbers of long-term unemployed over the last five years. Figure 2 shows how the structure of unemployment has changed over the last five years (for those aged 16 to 64). The initial increase in unemployment (up to mid 2009) was accounted for by the growth in short spells, as more individuals joined the unemployment rolls. Since then, there has been an increase in longer-term unemployment. The solid line (right-hand axis) shows those unemployed for more than a year as a proportion of the total number unemployed. This has increased by more than ten percentage points since the first quarter of 2008. The very long-term unemployed (dashed line) now account for nearly one in five of all unemployed.
These trends are worrying. Were the headline unemployment rate driven primarily by more people having short spells, one might imagine that their brief experience of unemployment would cause little or no harm to their longer-term prospects. In reality, we can see that there is a growing core of long-term unemployed. The concern is that their prolonged worklessness may make it increasingly difficult for them to return to employment or, in some cases, to find their first job. There are numerous possible reasons for this – employers may discriminate, morale may decline, skills may become obsolete, and so on. Whatever the reason, the personal reality for those who remain unemployed for too long is that there is a danger of them losing their ability to find a good job, or perhaps any job.
Figure 2. The changing structure of unemployment in the UK
Understanding the overall impacts – employment entry or employment retention?
From this perspective, the demonstrated ability of the UK Employment Retention and Advancement programme to increase employment rates among the long-term unemployment is of clear interest. Furthermore, Hendra et al. (2011) provide a detailed analysis to show that it was a net benefit, both to individuals and the government. What these results cannot tell us directly is whether these employment impacts were due to increased rates of employment entry or to an increased tendency to remain in work. A very small number of existing studies distinguish between these two effects. Card and Hyslop (2005) attribute the overall effect of the Self-Sufficiency Project primarily to faster exits from welfare, with only one-quarter due to employment retention. In Texas, Dorsett et al. (2013) also find that employment-entry effects of ERA dominate in Corpus Christi, with retention effects fading over time (as already noted, there were no sustained impacts in Fort Worth).
A new National Institute of Economic and Social Research discussion paper considers the case of UK ERA (Dorsett 2013). It represents the first evidence of its kind for the UK but, more generally, it adds to the evidence base by providing results for a very different population. Whereas both SSP and Texas ERA primarily targeted lone mothers on welfare, our results relate to the long-term unemployed – who face very different challenges to employment.
Key findings of the research
The key findings of the research are summarised in the charts below. Using the results from an econometric model of transitions into and out of employment, the overall impacts of ERA can be simulated. The top chart (labelled ‘Entry and retention effects’) simulates the overall effect of ERA. It shows an increase of about two percentage points in the employment rate – small in absolute terms, but 10% higher than in the control group. The confidence intervals (represented by dashed lines) indicate that these impacts are statistically significant. The bottom charts repeat this exercise, but this time suppress in turn the employment retention effects and the employment entry effects. Suppressing the retention effect allows the separate contribution of the entry effect to be seen (bottom left chart). From this it is clear that the impacts over the first three years or so were driven mainly by more people entering work, but that employment entry effects faded beyond that point. Suppressing instead the employment entry effect allows the separate contribution of employment retention to be seen (bottom right chart). After having no significant effect for the first three years or so, retention effects subsequently kicked in and became the main driver of the longer-term overall impacts.
Figure 3. The effect of UK ERA on employment by month since randomisation
The findings of this study are encouraging to the extent that they demonstrate the potential for temporary in-work support to help the long-term unemployed achieve a sustained move into work. Through increased employment retention, there is the hope of longer-term benefits such as increased employment stability, skill acquisition, earnings growth, career advancement etc., and that individuals can escape the ‘low-pay, no-pay’ cycle. The fact that this has been possible among a hard-to-help group such as the long-term unemployed is particularly encouraging.
When considering the policy relevance of these results, attention must be given to the fact that, since June 2011, back-to-work programmes in Britain have been streamlined into a single Work Programme. Under the Work Programme, support is provided – or at least coordinated – largely by private-sector companies who are paid according to the outcomes they achieve. These are structured to reward sustained employment rather than job entry. Given this, the findings reported here are of direct relevance. However, providers themselves are unlikely to introduce an ERA-type programme for two reasons. First, outcome payments are based on a two-year period of working with customers, which does not allow sufficient time for the additional income arising from longer-term impacts to outweigh the costs of operating the programme (which fall most heavily in the early years). Second, much of the benefit from ERA arises from increased taxes and reduced transfer payments – these do not accrue to providers, and consequently do not feature in their decision-making.
So, while ERA has been shown to be effective in increasing employment overall and employment retention in particular, there is little incentive for the private companies who deliver back-to-work services in Britain to introduce a similar type of support.
Card, D and D R Hyslop (2005), “Estimating the effects of a time-limited earnings subsidy for welfare-leavers”, Econometrica 73(6): 1723–1770.
Dorsett, R (2013), “The effect of temporary in-work support on employment retention: evidence from a field experiment”, NIESR Discussion Paper 411.
Dorsett, R, R Hendra, P K Robins, and S Williams (2013), “Can post-employment services combined with financial incentives improve employment retention for welfare recipients? evidence from the Texas employment retention and advancement evaluation”, NIESR Discussion Paper 409.
Hendra, R, K-N Dillman, G Hamilton, E Lundquist, K Martinson, M Wavelet, A Hill, and S Williams (2010), “How effective are different approaches aiming to increase employment retention and advancement?”, New York: MDRC.
Hendra, R, J A Riccio, R Dorsett, D H Greenberg, G Knight, J Phillips, P K Robins, S Vegeris, J Walter, A Hill et al. (2011), “Breaking the low-pay, no-pay cycle: Final evidence from the UK Employment Retention and Advancement (ERA) demonstration”, Department for Work and Pensions Research Report 765.
Michalopoulos, C (2005), “Does Making Work Pay Still Pay? An Update on the Effects of Four Earnings Supplement Programs on Employment, Earnings, and Income”, New York: MDRC.
Robins, P K, C Michalopoulos, and K Foley (2008), “Are two carrots better than one? the effects of adding employment services to financial incentive programs for welfare recipients”, Industrial and Labor Relations Review: 410–423.