Making benefits work

Spotlight on Employment
OECD Directorate for Employment, Labour and Social Affairs

Illustration by David Rooney

Attaching conditions to receiving benefits can prise people off dependency and attach them to the workforce. As many people on benefits can and should work, more should be done to encourage these “activation programmes” and help them perform better. 

We are getting older, but are we also getting sicker or less employable? Usually not, though a look at the trends does give rise to just such a question. In many OECD countries, the share of the working-age population receiving income-replacement benefits continued to increase in the 1990s, particularly for old age, disability, lone-parent and social assistance benefits. In some countries, most of the working age people who are neither employed nor studying receive some kind of income-replacement benefit.

There are some signs that the upward trend is levelling off. It has gone into reverse for unemployment benefits in some countries, such as Denmark and the UK. And following the introduction of “welfare-to-work” policies in the US, the number of adult welfare recipients is now three times lower than a decade ago. Still, benefit receipt without conditions attached remains widespread.

There are eight main categories of social protection benefits: old age; survivors (widows and orphans) – this is the only type of benefit to have declined in recipiency across the OECD since 1980; sickness; disability; maternity and home parenting; care benefit and other leave (e.g., sabbaticals); unemployment; and social assistance (including lone-parent benefits).

Overall benefit dependency varies from 11% of the working age population (15-64) in Japan and Spain – Spain however has one of the highest rates of people with no benefit but also no work – to 24% in France and 38% in the Slovak Republic – which has one of the lowest no-work/no-benefit rates. The largest overall benefit category in 1999 was disability (4.6%) followed by unemployment (4.1%). Parental/maternity (0.8%) and sabbatical-type leave (0.2%) were the smallest. Old age benefits (below 65) range from below 1% to over 7%, depending on the country. These benefits are a vital pillar of social welfare protection in all countries. But while not all people on benefits can work, many of them can and should. If benefits keep people out of work, that means a drain of resources away from other needs, with fewer people at work funding more people on support. What can be done?

Cutting benefit levels is one way, but this can cause people simply either to transfer to other available benefits or quit the labour force altogether and become even poorer. That is why policies based on “activation” programmes that get employable people off dependency and into work have become more popular.

There are many examples to follow. Much-praised egalitarian countries, like Sweden, are noted for the strictness of their criteria for receiving unemployment benefits. “The duty to work” has always been a core principle in Swedish policy, and unemployed people there are seldom inactive for long before they are assigned to a training or job creation programme. The system has recently been undermined by a “carousel” effect where people move between benefit and other public programmes without taking a regular job. Action is now being taken to limit this problem.

Even in the less-egalitarian US, when many lone parents were dissuaded from claiming benefits as welfare-to-work reforms were implemented, few of them found really good jobs. Nevertheless, outcomes were better than pessimists had feared. In fact, some two thirds of those that lost benefits are now in work, and average child poverty rates have fallen.

It is clear that for activation programmes to work, benefit eligibility criteria are important. Yet many types of benefits, like those for early retirement, as well as some disability and lone-parent benefits, do not require able people to be ready to work. Where possible, this should be changed. One attraction of activation measures is their value for money. Total spending on active labour market programmes ranges from 0.1% of GDP in Mexico and 0.2% in the US to some 1.7% of GDP in the Netherlands and 1.6% in Denmark. But this is only one measure. Of the 25 countries surveyed in the 2003 OECD Employment Outlook with above average spending in GDP terms, only Ireland, Norway and Sweden also spent sharply more than average on active unemployment schemes compared with inactive or passive (unemployment benefit) schemes. Denmark spent just 32% of its total unemployment outlay on active programmes in 2000, whereas Ireland and Sweden spent around 60% each. So while Denmark invests generously on active schemes, it pours more money into passive unemployment as well. Although this spending may be important, many activation measures available incur relatively little in extra budgetary costs.

Tightening benefit eligibility while applying activation principles can help reduce beneficiary numbers and raise employment. But simply exerting pressure on unemployed people to take jobs more quickly can edge people out of the workforce. Where no benefits are available, or if the conditions for claiming them become too strict, young people and married women in particular will just turn to other family income for support instead. These risks can be reduced by keeping the level of benefits high enough to discourage people from slipping out of the labour force.

Providing job seekers with latitude can also help. In the US state of Maryland, experimental groups that had to search for work were released from the routine requirement of reporting back every week. Their average duration of unemployment increased slightly, but so too did their total annual earnings on finding a job, by some 4% compared with groups that had to report back. The freed-up job seekers held out longer for the right job, or searched more efficiently.

In contrast, a UK study of the Job Seeker’s allowance found that after tighter benefit legislation, male job seekers’ mean earnings on re-entering work fell by a fifth. However, many other studies have found that activation measures succeed in getting people into work without loss of job quality. In the US National Evaluation of Welfare to Work Strategies, most of the experimental programmes examined had a positive impact on earnings, and for several of them, for as much as five years after participation in the programme. Intensive employment counselling – teaching job seekers to search out the right kind of job for them – can even lift earnings. The trick is to combine pressure to take up work with good quality assistance, as well as training.

Of all the benefit types, one of the hardest nuts of all to crack is disability: this type of welfare has risen in the OECD, particularly in Canada, Ireland, the Netherlands, Sweden and the UK. Yet few, if any, indicators around suggest that OECD people’s health has declined. But to judge from past trends, it seems that when more people are paid benefits on the basis of complaints like mental and physical stress or bad backs, more people seem to report similar problems. With the right policies, such people can often be kept in work, even part-time.

Activation policies can work well. But their impact is greatest where labour markets function smoothly and job openings are continuously announced. If very few jobs are available, say, because hiring costs are too high or recruitment rules too tight, activation initiatives, no matter how well designed, will be blunted. This would lead to a loss not only of political support, but of public motivation as well.


OECD (2003), Employment Outlook, Chapter 4 “Benefits and employment, friend or foe?”, Paris.

OECD (2003), Transforming Disability into Ability: Policies to Promote Work and Income Security for Disabled People, Paris.

OECD (2001), Labour Market Policies and the Public Employment Service, Paris.

©OECD Observer No 239, September 2003

Economic data

GDP growth: -9.8% Q2/Q1 2020 2020
Consumer price inflation: 1.3% Sep 2020 annual
Trade (G20): -17.7% exp, -16.7% imp, Q2/Q1 2020
Unemployment: 7.3% Sep 2020
Last update: 10 Nov 2020

OECD Observer Newsletter

Stay up-to-date with the latest news from the OECD by signing up for our e-newsletter :

Twitter feed

Digital Editions

Don't miss

Most Popular Articles

NOTE: All signed articles in the OECD Observer express the opinions of the authors
and do not necessarily represent the official views of OECD member countries.

All rights reserved. OECD 2020