VoxEU Column Global crisis Labour Markets Welfare state and social Europe

Welfare to work: Sticks rather than carrots

Rising unemployment has forced policymakers to look for ways to get the unemployed back to work – to raise the “reemployment” rate of the unemployed. This column provides new evidence from the Netherlands suggesting that the stick of benefit sanctions is much more effective than the carrot of reemployment bonuses.

Given difficult circumstances, governments are considering various policy instruments to increase the “reemployment” rate of unemployed workers. This is no easy task – traditional active labour market policies are often not very successful.

  • Card et al. (2009) find that subsidised public sector employment programmes are relatively ineffective (based on approximately 200 European and US microeconometric evaluation studies). 
  • Job search assistance programmes have a favourable impact, especially in the short run, while classroom and on-the-job training programmes are not favourable in the short run, but have more positive impacts after two years.

This suggests that, in the short run, the programmes that work best are those that stimulate job search and that have clear incentives for the unemployed. This is not surprising. For example, on the basis of outcomes from four US cash bonus experiments and six job search experiments, Meyer (1995) concludes that economic incentives affect the speed by which unemployed workers leave unemployment insurance. Some of these job search experiments included enforcement of job search rules and administrative measures of enforcement such as denial of benefits.

But what about the long-term effects? In this column, we consider long-term welfare benefits recipients in the Netherlands rather than short-term unemployment insurance recipients in the US.

New evidence: Welfare recipients in the Netherlands

In the Netherlands, welfare benefits are a safety net and provide income support to unemployed workers who are not entitled to any other social insurance benefits like unemployment insurance or disability insurance benefits.

Welfare benefits recipients are often long-term unemployed with poor labour market prospects. Welfare benefits are means tested and related to the family situation, but not limited in duration. Benefit replacement rates are usually somewhat higher than in most other European countries and much higher than in the US (Van den Berg, Van der Klaauw and Van Ours, 2004).

Welfare recipients have an obligation to actively search for work. However, the administration of welfare is at the level of the municipality. Van der Klaauw and Van Ours (2010) investigate how positive and negative financial incentives affect the outflow from welfare to work. They exploit that during some period welfare recipients in Rotterdam were exposed to both financial carrots as well as financial sticks, i.e. the incentives could have positive or negative effects on the income of benefit recipients.

Reemployment bonuses and benefit sanctions

In 1997 reemployment bonuses were introduced in Rotterdam. Although the level of the benefits and some details for eligibility changed, until 2001 the main condition was that the individual should be on welfare benefits for at least one year. Furthermore, employment should last for at least six months to be entitled for a first payment.
In 2002 the cash bonuses specific for Rotterdam were replaced by a nationwide additional tax credit with roughly the same eligibility conditions. In 2003 all reemployment bonuses were abolished.

There is a fundamental difference between the US experiments and the setup of the Rotterdam bonus system. Whereas in the US experiments a bonus was paid to workers who found a job quickly, in Rotterdam welfare recipients became entitled for a reemployment bonus after being unemployed for at least one year. The main reason for this was fear of dead weight loss, i.e. workers taking up bonuses who would have found a job anyway. On the other hand, by focusing on long-term benefits recipients, one might stimulate individuals on welfare for almost one year to postpone starting work.

The Dutch welfare system also incorporates the possibility to impose punitive reductions on the benefit. Such benefits sanctions are often imposed for noncompliance with job search guidelines. Other reasons are because of administrative infringements like returning late from holiday, filling in forms incorrectly, etc. The nationwide government sets the rules and procedures for the benefits sanctions, but again municipalities implement the rules. Benefit sanctions are most often a 5%, 10%, or 20% reduction of the benefit during one or two months. But in case of serious fraud benefits may be much more severe.

Our analysis

In recent research (Van der Klaauw and Van Ours 2010) we analyse micro data from 2000 until 2003, covering the period with cash bonuses, tax credits, and no bonuses. The data contain 28,039 individuals between 16 and 60 years old who started collecting welfare benefits in Rotterdam. In total, the data include 30,527 welfare spells. There are thus some individuals with multiple spells, but the vast majority of the individuals only experienced a single period of receiving welfare benefits within the observation period.

Using a continuous-time duration model, we find that the effects of promising reemployment bonus are small and insignificant. This holds both for the cash bonuses and the tax credits. Benefit sanctions, however, have a positive and significant effect on the reemployment rate of welfare recipients.

To get some insight into the size of the effects we used our estimated model to perform some simulations focusing on job finding within two years after entering welfare. Table 1 shows the results of these simulations.

Table 1. Model simulation for the job-finding rate within two years after entering welfare (percentages)

  Males Females
No bonuses and no sanction 62.5 48.9
Bonus but no sanction 62.9 49.9
No bonus but sanction after one year 65.4 54.9
Bonus and sanction after one year 65.9 56.2


Source: Van der Klaauw and Van Ours (2010)

  • Without reemployment bonuses or sanctions, about 48.9% of the women and 62.5% of the men starting collecting welfare benefits find work within two years.
  • The reemployment bonus scheme only slightly increases this to 49.9% and 62.9% for women and men, respectively.

These effects are small, but in line with previous reemployment bonus studies that also find small effects.

  • Imposing a sanction after one year of welfare causes a much more substantial increase in the reemployment rate. The reemployment rates increases to 54.9% and 65.4% for women and men respectively. 
  • Finally, the two financial incentives combined increase reemployment from 48.9% to 56.2% for women and from 62.5% to 65.9% for men.
Sticks work better than carrots

The results suggest that benefit sanctions are much more effective in stimulating the transition from welfare to work than reemployment bonuses. A possible explanation is that benefit sanctions include an immediate financial consequence for the welfare recipient, while reemployment bonuses are a promise on a payment in the future. Welfare recipients may suffer from some degree of present bias and may heavily discount uncertain delayed payments such as cash bonuses. Furthermore, it should be noted that the take-up rate of the cash bonuses is only 38%.


Card, D, J Kluve, and A Weber (2009), “Active Labour Market Policy Evaluations: a Meta-Analysis”, IZA Working Paper 4002, Bonn.
Meyer, BD (1995), “Lessons from the U.S. Unemployment Insurance Experiments”, Journal of Economic Literature, 33:91-131.
Van den Berg, GJ, B van der Klaauw, and JC van Ours (2004), “Punitive Sanctions and the Transition Rate from Welfare to Work”, Journal of Labour Economics, 22:211-241.
Van der Klaauw, B and JC van Ours (2010), “Carrot and Stick: How Reemployment Bonuses and Benefit Sanctions Affect Job Finding Rates”, CEPR Discussion Paper 7924.