1The current Dutch disability insurance scheme (WAO) is heading for abolition in 2006. The scheme was introduced in 1967 as a unique approach to covering earnings loss due to long-term disablement. It was not only unique in terms of generosity and accessibility, but also in the broadness of its risk definition. The WAO integrated two pre-existing schemes: one covering work injuries and occupational diseases, and one covering other causes of disability. Like everywhere else, the work injury scheme was more generous, had a well-defined framework of loss categories, and offered immediate full coverage. And the definition of covered risk was limited and unambiguous.
2The unique step the Dutch took in 1967 was to broaden the work injury scheme to cover all disability contingencies, whether work-related or not. Its special features proved its weaknesses because it combined the usual generosity of a work injury scheme with a broad risk definition, including a wide range of non-specific, subjective health complaints.
3In this paper we document how the Dutch tried to cope with this unmanageable disability programme. The most important reforms are summarised in Table 1 (see appendix) As a background, we start by describing the vocationally handicapped population. Almost 16 per cent of the Dutch working-age population report having an impairment that limits the amount or kind of work they can do. About half of the 1.6 million vocationally handicapped are recipients of a disability benefit. This paper deals with both groups and gives some institutional detail on benefit entitlements and in kind provisions. Section 3 reviews the current sickness and disability benefit programmes. Section 4 discusses trends in disability expenditures and beneficiaries and documents the interaction between institutional changes and disability insurance claims. In section 5, we give an overview of the changes that took place over the past ten years and their efficacy. One of the changes was the introduction of a vocational rehabilitation infrastructure. At the end of this section we discuss proposals to replace the current programme by one that is less generous but also less contentious and more sustainable. Section 6 concludes by outlining some lessons to be learned from the Dutch disability experience.
A snapshot of the Dutch vocationally handicapped population
4Survey data from 2003 show that 15.8 per cent of 10.9 million Dutch citizens of working age (15-64) report having (1) a chronic impairment, and (2) one that restricts their work capacity. This group is called “Arbeidsgehandicapten” (vocationally handicapped) . Compared to this average, the prevalence of handicaps is relatively high among women (17.3 per cent) and in the age group 55-64 (28.3 per cent).
5While this prevalence rate is 20 per cent among those with a low level of schooling, it is only 10 per cent among the best schooled. The presence of vocational handicaps is more frequent among immigrants from non-Western countries (20 per cent), and specifically among those of Turkish or Moroccan descent (25 per cent).
6The labour force participation rate of the vocationally handicapped is 48 per cent. This is 21 percentage points lower than the national average of 69 per cent. Moreover, their unemployment rate was 8.2 per cent, while the average unemployment rate was 5.3 per cent in 2003.
7The vocationally handicapped represent 10.7 per cent of the employed population. There is no mandatory quota rule for the employment of handicapped workers in the Netherlands but in countries where such a rule exists the quota is usually lower than 10 per cent. However, while on average an unusually high 36 per cent of Dutch workers work part-time, 44 per cent of handicapped workers do so. Not only do they more often work fewer hours, but also 38 per cent have lower-scaled jobs, against an average of 30 per cent.
8These data confirm that the vocationally handicapped work less and, if they work, that they earn less that the average worker. In other words, many handicapped workers are also handicapped financially.
The current Dutch sickness and disability schemes
9The Dutch Civil Code stipulates that employers are obliged to pay 70 per cent of gross wage earnings if an employee is unable to perform his or her job because of illness or injury, irrespective of its cause. Under collective bargaining agreements between employers and employees, sickness benefits are supplemented up to the level of net earnings. As of 2004, sick pay ends after 24months. Before 2004, sick pay would run out after 12months.
10Since March 1996 employers are fully responsible for financing sick pay. They may reinsure their sick-pay liability with a private insurer but they are not obliged to do so. Before 1996 sectoral funds paid sickness benefits. These funds were fed by sector-specific premium rates. Hence, collective coverage was replaced by mandating the individual employer to cover sick pay.
11Employers are also mandated to contract with a private provider of occupational health services to manage absenteeism. Doctors employed by these occupational health agencies check whether the absence from work is legitimate and give a prognosis concerning work resumption.
12Small firms may be unable to offer a commensurate job if an employee acquires a disability that prevents him from doing his or her job. In that case, a reintegration service provider should mediate towards placement in a new firm. As of 2003 employers are obliged to subscribe to the services of a private reintegration firm to help disabled employees for whom no commensurate work is available within the firm to find new employment.
Hence, privatisation of sickness benefit payment went alongside privatisation of administration and management of sickness absenteeism.
13Under the Dutch ruling, any illness or injury entitles an insured person to a disability benefit after a mandatory waiting period of 24 months. While other OECD countries make a distinction as to whether the impairment occurred on the job or elsewhere, only the consequence of impairment is relevant for the Dutch disability insurance programme.
14Two separate benefit programmes targeting different social groups provide compensation for loss of earning capacity due to long-term or permanent disablement. The first, and by far the biggest, programme covers employees, and awards wage-related benefits. The other one addresses those handicapped from youth. This scheme provides flat benefits at the social minimum level. These are financed out of general revenue. Eligible handicapped young people are entitled to a benefit from age 18 onwards. Otherwise, the design and administration of this programme is the same as the wage-related programme. 
15The degree of disablement is assessed by consideration of the disabled worker’s residual earning capacity. Capacity is defined by the earnings resulting from any job commensurate with a person’s residual capabilities as a percentage of earnings, irrespective of education and work history. The degree of disablement is the complement of the residual earning capacity and defines the benefit level. The Disability Insurance programme for employees has seven disability classes. The minimum loss of earning capacity entitling a person to a benefit is 15 per cent. Wage replacement rates range from 14 per cent of covered earnings in the 15 to 25 per cent disablement category to 70 per cent in the 80 to 100 per cent category.
16The disability scheme for handicapped young people has six disability categories: it skips the first category so that entitlement starts at a degree of disability of 25 per cent. The earnings base for calculating benefits is the minimum wage. Hence, the benefit at full disablement is 70 per cent of the minimum wage. 
17Partial benefits can be combined with labour earnings up to the level of the pre-disability wage. If recipients of a partial benefit are unable to find gainful employment they are entitled to partial unemployment benefit.  Combination of disability and unemployment benefits never replaces more than 70 per cent of earnings lost.
18Wage-related benefits are based on age and earnings. The disability benefit period is divided into two, chronologically linked parts. The first is a short-term wage-related benefit replacing 70 per cent of pre-tax earnings. The duration of this wage-related benefit depends on age at the onset of disablement. It varies from zero for those under age 33 to six years for those whose disability started at age 58 or over. Hence, workers age 58 and over keep their 70 per cent replacement rate until the statutory pension age 65. For older workers the accrual of pension rights related to their last job continues after entering the disability rolls. In addition, most pension plans do not require disability beneficiaries to pay pension premiums. Such contract rules discourage re-entry into the labour market by creating a gap in pension accrual rights, and make the disability system an alternative early retirement option. 
19The second part is a so-called follow-up benefit with a lower income base and hence a lower replacement rate with respect to the pre-disability wage.
During the follow-up period, the earnings base for benefit calculation is the minimum wage plus a supplement depending on age at onset, according to the formula 2.0 per cent x [age at onset -15] x [wage –minimum wage].Age serves as a proxy for work history, or “insurance years”, introducing a quasi-pension element into the disability system. Most collective bargaining agreements cover the gap between the lower replacement rates in the follow-up period and the 70 per cent replacement rate during the first period of disablement. The effective replacement rate when fully disabled, therefore, stays at 70 per cent in most cases.
Disability benefits are capped by a maximum amount of covered earnings equalling €43,770 per annum. This is also the maximum amount of income taxable for disability (and unemployment) insurance.
Trends and issues
20In May 2002 the British weekly The Economist commented on the Dutch economy in an opinion article titled “Going Dutch”: “ (…) it is the very need for consensus that has inhibited further reforms to the much-abused and excessively generous disability system, which pays out to a ludicrous one in seven Dutch people of working age.” In the eyes of this commentator the Dutch disability experience is a clear illustration of the negative side of the much-praised culture of consensus and tolerance in Holland.
What happened really?
21The data collected as part of an OECD disability policy project  both confirms and refutes the stereotype of the Dutch disability system referred to in this quote. According to Table 2.1 of the OECD report, Holland is still among the big spenders of disability benefits but it is not the biggest spending country anymore, as it was in 1990. In 1999, broad disability benefit expenditures were 4.14 per cent of GDP, which was 28 per cent lower than in 1991. 
22Figure 1 shows the trend in the number of persons receiving a disability benefit as a percentage of the labour force (including disability beneficiaries),  and disability benefit expenditures as a percentage of GDP. Disability benefits are here defined in a narrow sense, including both benefits from contributory and non-contributory disability schemes. From a 1985 high of 4.2 per cent of GDP disability benefit, expenditures decreased to 2.6 per cent in 2003.
DI beneficiaries as % of the labour force and DI benefits as % of GDP, 1971-2003
DI beneficiaries as % of the labour force and DI benefits as % of GDP, 1971-2003
23At the same time, however, the relative number of beneficiaries stayed at 10 per cent of the labour force – the level it had reached in 1981, after the disability explosion of the 1970s. In absolute terms, the number of disability beneficiaries grew continuously from 475,000 in 1976 to 921,000 in 1993.  Changes in the definition of disability and in the way benefits are calculated drastically reduced the number of new awards. Moreover, the status of some of the current beneficiaries was reviewed using the new, more stringent definition. This increased the number of benefit terminations and led to a 7 per cent drop in the number of beneficiaries to 855,000 in 1996. From then on the numbers started growing again and reached 979,000 in November 2002, coming close to the politically contentious level of one million disabled.
24The Figure above shows that the reduction in spending on disability benefits was not caused by a smaller number of beneficiaries. The average benefit must therefore have gone down. Over the 30 years covered by Figure 1, cutting statutory benefits appeared to be the only policy measure to reduce the financial burden of an otherwise uncontrollable programme. In the early 1980s, beneficiaries lost 25 per cent of their purchasing power through a series of substantial retrenchments. First, levying social insurance contributions on benefit income changed the calculation of after-tax benefit amounts. In 1982 and 1983, the after-tax DI-benefit level was reduced through the abolition of certain tax exemptions for the disabled. In 1984, the earnings base from which benefits were calculated was reduced. Moreover, all incomes under government control – transfers, civil servant salaries, and the statutory minimum wage – suffered a 3 per cent nominal cut. Finally, in 1985, replacement rates (before tax) were lowered from 80 to 70 per cent of last earnings, when fully disabled. These direct cuts were accompanied by the elimination of the system of automatic indexation (adjustment) of government-controlled incomes. Benefits were cut again in August 1993, when statutory replacement rates were reduced according to age at the onset of disability. As a result benefits lost another 20 per cent of their real value between 1985 and 1995. This loss contrasts sharply with per capita GDP, which increased by one-third during the same period. To summarise, the after-tax replacement rate for an average worker who becomes fully disabled dropped from about 90 per cent in the 1967-1980 period to 85 per cent between 1980 and 1994, and again to 75 per cent from 1994 onwards.
25After the changes of 1993 the share of partial benefits grew sharply. As a result of these changes, the notion of suitable work was eliminated from the definition of disability. Capacity is since defined by the earnings resulting from any job commensurate with a person’s residual capabilities as a percentage of pre-disability usual earnings. The degree of disablement is the complement of the residual earning capacity and defines the benefit level. Before 1994, only jobs that were compatible with a person’s training and work history could be taken into consideration in the assessment of residual capacity. This new ruling made the percentage of partials among new awards grow from 19 per cent in 1990 to 45 per cent in 2001.
26Two-thirds of partial-benefit recipients work. For them, and their employers, the benefit acts as a wage subsidy. Research has shown that partial beneficiaries differ from full beneficiaries in many respects: they are older, better schooled, more often male, married and the main breadwinner, have a longer tenure with their current employer and work in large, financially healthy firms.  In short, Dutch partial beneficiaries are socially and economically better off. The data suggest that partial benefits often are used to offer older employees easier work conditions and act as a partial early retirement scheme.
The profile of the average beneficiary has changed
27Over the past three decades the typical new disability beneficiary changed from an older male industry worker with a long work record in physically strenuous work into a younger female employee in the service industry with a relatively short labour-market record. As 57 per cent of Dutch women work part-time, their wages and their D.I. benefits are lower.  An increasing proportion of women among D.I. entrants, therefore, implies lower benefits, other things equal.
28Figure 2 displays the disability beneficiary incidence rate for men and women. Women had lower rates until 1985, and have had higher ones ever since. More importantly, the gap between the two incidence rates increased continuously from 1983, when the female rate was 15 per cent lower than that of men, until 1998 when women had an 80 per cent higher chance of becoming dependent on disability benefits. It has stayed at that level since.
New beneficiaries as a % of the labour force, by gender, 1971-2003
New beneficiaries as a % of the labour force, by gender, 1971-2003
29In absolute terms the total number of disability beneficiaries increased by 6 per cent between 1991 and 2001. While the male beneficiary volume decreased by 13 per cent, its female counterpart increased by 43 per cent.
30The sharp increase in female disability was matched by an equally strong growth in labour-force participation of mothers. Traditionally, the Netherlands had very low labour-force participation rates among mothers. In the 1970s, three out four women stopped working after the birth of their first child. Twenty years later, only one-third of women stops working. In other words, the traditional single-earner model was replaced by one in which husbands work full-time and wives have part-time jobs. As a consequence of this social change, the lack of sufficient childcare facilities has been compensated for by the disability scheme. Disability benefits allowed market production be replaced by home production without a sharp drop in household income. The feminisation of disability benefit dependency illustrates how an income-oriented disability policy invites the strains of dual earnership to be defined in medical terms.
Reforms after 1990: privatisation, reintegration, and a new bill
31In the early 1990s, Dutch policymakers turned to defining the disability issue in terms of “moral hazard”. They concluded that the system lacked appropriate incentives for the three parties directly involved: employees, employers and system administrators.
… for employees
32Employees were hit by benefit cuts in 1993, when the “two-phase” system was introduced – a wage-replacement phase followed by a phase with a lower, age-dependent replacement rate (see section 3). Although collective-bargaining agreements corrected the gap between the lower rate in the second phase and that in the first phase, for most employees these supplements replaced supplements at the standard 70 per cent replacement rate up to 100 per cent of the net wage. All in all, the effective rate went down for most employees.
33As part of the 1993 amendments the definition of disability under the Disability Insurance Act became stricter. The loss of earning capacity used to be assessed against work that was considered suitable with respect to a person’s level of education and achieved level of functioning. If a disability prevented employment in suitable work, a person was considered fully disabled. As of August 1993, the extent of disablement is assessed by considering the complete labour market, instead of the parts considered suitable. This is one of the reasons why the share of partial benefits increased sharply. Moreover, the disability status of all beneficiaries under age 45 was reviewed according to the new standards. These reviews led to a surge in terminations and reductions of benefits.
34The 1993 amendments also affected the incidence of new disability awards (see Figure 2). The decrease in awards and loss of part of their financial appeal may well be the combined result of more stringent eligibility requirements and lower application rates. But the benefit cuts, together with a booming economy and an increasingly tight labour market during the last six years, also changed the face of disability. Better-paid workers with long careers tried to avoid becoming fully dependent on disability benefits as they had better opportunities in the labour market. Secondary breadwinners and low-wage workers lose less when going on disability and are therefore strongly over-represented among both the disabled (see section 2) and disability benefit recipients.
A smaller number of awards and a steep increase in benefit terminations resulted in a 7 per cent decrease of the disability beneficiary population over the three-year period between 1994 and 1996. Until 2003, these were the only years in which the number of beneficiaries declined since the introduction of the comprehensive disability insurance scheme in 1967.
… and for employers
Privatisation of sickness benefits
35In March 1996, the Sickness Benefit Act was abolished. Under this act, sick pay was collectively financed through sector-specific insurance funds. As mentioned in section 3, abolishing this act meant that employers became responsible for coverage of sick pay during the first 12 months of sickness. As of 2004 this self-insurance period is extended to 24 months.
36Up until now, firms are legally mandated to contract with a private occupational health agency and buy a prescribed package of services including prevention of and monitoring periods of sickness.  These legal mandates sought to reduce absenteeism and inflow into the disability benefit programme by confronting firms with the full cost of employee absence. It made investment in prevention and reduction of sickness profitable as it reduced avoidable costs of absenteeism.
37Sickness absence rates dropped from 8 per cent in 1990 to 6 per cent in 2000 – a 25 per cent drop.  Both these years represent a cyclical high and their comparison therefore shows the influence of the business cycle on absenteeism. At least part of this large drop can be ascribed to privatisation, and its associated incentives. This favourable result is obtained despite the fact that about 80 per cent of all firms took out some form of private insurance to cover their sickness liabilities.
38There appears to be a strong negative relationship between firm size and insurance coverage: while firms with fewer than 20 employees have a coverage rate of about 83 per cent, only 25 per cent of those with 100 or more workers buy insurance. Larger firms also choose a larger coinsurance period or buy a stop-loss arrangement.  To avoid adverse selection, insurance companies stipulate that no employee be excluded from coverage under a sick-pay insurance policy purchased by the employer. Insurers also demand that firms contract occupational health agencies, and stipulate which set of services is to be contracted. Econometric analysis shows that the insurance status of a firm has no adverse effect on its consecutive absence record.  This implies that apart from experience rating, insurers may use other instruments to control damages.
Surprisingly, privatisation did not induce a surge in conflicts between sick workers and employers refusing to continue paying wages. This may have been the result of the fact that the privatisation took place in a boom period. But also the 2000-2004 recession did not lead to significant trouble.
Disability contribution rates are experience-rated
39Since 1998, experience rating of firms has been gradually phased into the disability insurance scheme. Pre-1998 benefits are still funded by the existing uniform pay-as-you-go contribution rates but as of 1998 the first five years of disability benefit recipiency awarded to new beneficiaries is paid out of premiums that are levied according to the “polluter pays principle.” If an employee is awarded a disability benefit, the firm will face a higher contribution rate, and vice versa if a firm employs a disability beneficiary. Moreover, firms are allowed to opt out of the public insurance system, but only with respect to the coverage of the first five years of benefit recipiency.
40The disability insurance scheme for employees is financed by levying two separate premium rates, both paid by the employer. The first is a uniform pay-as-you-go rate covering the benefits of those that were already on the rolls before 1998. As of five years after its start – i.e. from 2003 – the pay-as-you-go rate also covers benefits that commenced after 1997 and are awarded for more than five years. Over the past five years this rate has gone down from 7.55 per cent to 5.05 per cent of the taxable wage base (up to €43,000 per year).
41The second rate covers the first five years of benefit recipiency and is differentiated according to the firm’s specific disability risk. To calculate the risk in year t, the total expenditures on disability benefits for the firm’s disabled employees in year t-2 is expressed as a percentage of the average wage bill over the past five year period. The firm’s specific risk determines the differentiated rate. The average risk rate increased from 0.30 per cent in 1998 to 2.38 per cent in 2003. As of 2003 only firms with a wage bill of more than €600,000 pay differentiated rates. These rates are confined within lower and upper limits. The upper limit increased from 1.12 per cent in 1998 to 8.52 per cent in 2003.
42Five years after the inception of experience rating the system can be considered mature. The uniform pay-as-you-go rate is expected to stay at about the current level of 5 per cent of taxable wages. The evolution of the average risk rate strongly depends on how inflow rates develop.
Koning examined the effects of this system of experience rating, using a unique longitudinal data set consisting of the Dutch disability insurance (DI) administration records.  The data cover about 370,000 firms employing roughly six million insured workers. These firms were followed over a three-year period, from 2000 to 2002. The overall picture that emerges from his empirical analysis is that the impact of experience rating on DI inflow has been substantial. Employers increased their preventative activities in reaction to an increase in their premium rates (“ex post incentives”).
… but not for the gatekeepers of the programme
43In the debate on disability policy the focus gradually shifted from the programme itself toward the programme administrators. In 1993, a multi-party parliamentary committee investigated the operations of the then-existing Insurance Agencies which were organised by sector of industry and held a legally protected monopoly with regard to the administration of sickness, disability and unemployment insurance benefits. The committee devoted special attention to the administration of the disability insurance scheme. The committee publicly interrogated a vast number of current and former administrators, civil servants, and politicians. The picture that emerged from the nightly televised summaries was devastating for the image of the Insurance Agencies. What had been suspected, and what had already been shown by research, was now publicly confirmed. The committee’s report created broad political support for drastic changes regarding, in particular, the dominant, and autonomous, position of the trade unions and employers” representatives in the management of social insurance.
44In 1995, as a result of the committee’s recommendations, an independent supervisory body was set up. It publishes annual reports on the efficiency and legality of the administration of the social insurance programs. In 1997, the public Insurance Agencies that were run by the social partners were privatised, and regrouped themselves into five organisations. Next to their traditional tasks in administering public insurance programs (unemployment and disability), they set up a range of private activities, offering medical and vocational rehabilitation, and occupational health and employment services.
45The original plan was to create a competitive market on which these five agencies, as well as new entrants to this market, would compete for contracts with companies or groups of companies to administer wage-replacing unemployment and disability insurance. The trend was towards offering “full-service packages” that would cover the legally mandated social insurance liabilities as well as pensions, health insurance and outplacement of redundant employees.
46The public debate on this model of private provision of social insurance exposed several problems. A competitive insurance market for mandatory coverage of disability risks could be viable, and efficient, if private insurers were allowed to control all the links in the “insurance chain”: running from drafting policies, calculating premiums, administering indemnities, controlling damages and managing claims. Insurers could offer firms tailor-made packages by varying elements such as the extent of co-insurance and the intensity of damage control through prevention, swift rehabilitation and monitoring activities. One crucial element in this chain is the assessment of the degree of disablement. A political majority was unwilling to subject disability assessment to the business interest of private insurers. As a consequence, a hybrid model was proposed in which the whole chain was privatised except disability assessment, which was to be done by a separate public (medical) agency.
47Second, while disability is a privately insurable risk, unemployment is not. Privatisation of unemployment insurance would not obtain a (socially) efficient market. Apart from the insurmountable problem of risk dependency, employers would only be interested in the cheapest unemployment insurance administration contract because they would not profit from investment in quick re-employment of workers after they became redundant. Putting disability and unemployment risks in one basket, therefore, would result in a (socially) sub-optimal outcome.
48Third, private agencies that cover the mandatory (public) insurances are likely to offer additional, related, insurance services, such as health insurance and pensions. Insofar as the portability of such employee benefit packages is limited, employees are tied to the firm. Likewise, firms may find it difficult to change providers of employee benefits.
And, finally, private agencies that obtain data on covered workers because they run public schemes may abuse them for other commercial activities, e.g. risk selection for health insurance. Similarly, they may use moneys from mandatory, public, insurances for their private business. Auditing such hybrid organisations is complex, controversial, and expensive.
For these and other reasons a political majority pulled the plug on this privatisation plan in the summer of 1999. In 2002 the Social Insurance Institute was established to run the disability and unemployment insurance schemes as a so-called quango (quasi-autonomous non-governmental organisation) under contract with the Ministry of Social Affairs and Employment. Only rehabilitation (reintegration into paid work) is contracted out to private firms. This could offer an opportunity for the existing occupational health service companies that now do the management of sickness benefit claims to broaden their scope.
49In contrast to countries with a similarly broad social welfare system, the Dutch disability programme used to lack effective mandates regarding vocational rehabilitation and a rehabilitation infrastructure to support such mandates. This was increasingly felt as a system failure. In July 1998 the Act on Reintegration of Work Handicapped Persons (REA) introduced a new target group. Under this Act the diverse provisions in kind and subsidies that were previously to be found in a number of schemes were grouped together and made consistent.
50Work-handicapped persons are all those:
- that have a disability that reduces their productive capacity, and
- are entitled to a disability benefit, or those that have lost their entitlement less than five years ago;
- are entitled to an in-kind provision or subsidy to maintain or restore their productivity, or those that have lost their entitlement to such provision less than five years ago;
- belong to the group targeted by the Sheltered Work Provision Act;
- do not belong to any of the before-mentioned groups but have been assessed (through medical examination at a social insurance agency or by an occupational heath service) as being work-handicapped. For this group the work-disabled status is allowed for five years after which it has to be re-established.
51As of 2002, REA covers the following types of provisions:
- – Work-handicapped employees may be entitled to schooling, training, mobility provisions, trial placement and personal assistance, and certain therapies (such as stress and RSI training) to maintain or restore their productivity
- – Companies pay a lower disability insurance rate and are exempt from experience rating for handicapped workers. The sickness benefits of handicapped workers are covered collectively so that their employers do not bear the financial risk of continued wage payment in the case of sickness.
- – Companies are entitled to subsidies that cover the cost of adapting the work place for disabled employees.
- – In 2001, 57,000 REA provisions were awarded to employees and about 50,000 to employers. As employees, or their employers, often receive more than one provision the number of employees receiving a provision is much lower than the total number of provisions given in 2001 (107,000). But even if this figure corresponded to the number of workers receiving a REA provision, it is small compared with a target group of 800,000 non-working handicapped persons.
52This conclusion is based on a study of workers who are on long-term sickness but still have an employment contract with their current employer. In this case, the employer and its occupational health agency make the application for a REA provision. Concerning disability beneficiaries, REA provisions are usually part of a reintegration (back-to-work) plan drafted by vocational experts of the Social Insurance Institute. These plans may be compared with the “individual participation packages” proposed by the OECD. On the one hand beneficiaries can influence the design of the plan by stating their preferences for certain REA provisions and lines of work; on the other hand, beneficiaries are legally mandated to take all steps necessary to restore their productive capacities. Therefore, those that are offered a plan cannot refuse to co-operate, unless they can prove that they already are on the road back to work.
The Social Insurance Institute contracts with private reintegration service organisations to execute reintegration plans. This is done by parcelling out groups (plots) of beneficiaries to organisations with the best offer in terms of price, successful placement record and professionalism. These plans are financed out of the REA budget, and cover both the reintegration instruments and the effort of the reintegration service. In 2001, about 50,000 reintegration plans were contracted for the work-handicapped. About half of these concern disability beneficiaries; the others are unemployed disabled persons (with or without an unemployment transfer income). In other words, plans were made for about 2.5 per cent of the disability beneficiary population. Contractually, 35 per cent (17,500 of 50,000 plans) should result in successful placements (employment for at least six months).
53As of April 2002, the responsibilities of the sick employee, his/her employer, and the occupational health service are legally specified, and require a structured approach to early intervention in cases of sickness. After a maximum of six weeks of absence the occupational (health service) doctor has to make an initial assessment of medical cause and functional limitations and give a prognosis regarding work resumption. On the basis of these data, employer and employee together draft a vocational rehabilitation plan in which they specify an objective (resumption of current/other job under current/adapted conditions) and the steps needed to reach that goal. They appoint a case manager and fix dates at which the plan should be evaluated, and modified if necessary. The rehabilitation plan should be ready in the eighth week of sickness. It is binding for both parties, and one party may summon the other when considered negligent.
54After 35 weeks of sickness the Social Insurance Institute sends a Disability Insurance application form to the sick employee. Disability Insurance claims have to be delivered before the fortieth week of sickness. Claims are only considered admissible if they are accompanied by a rehabilitation report, containing the original rehabilitation plan and an assessment as to why the plan has not (yet) resulted in work resumption. If the report is delayed, incomplete or proves that the reintegration efforts were insufficient, the claim is not processed and the employer is obliged to continue paying sickness benefits even after the waiting period for disability benefit has elapsed.
55This is a serious step in the direction of mutuality of rights and responsibilities both in the relationship between employer and employee, and of both parties in their relationship with the state, represented by the Social Insurance Institute. Employees who consistently refuse to co-operate with their employer to execute the plan can be dismissed. To that end the labour law has been changed, because previously an absolute dismissal ban was in force for the first two years following the onset of a disability.  Employers can be sanctioned by a one-year extension of the payment period of sickness benefits if proven at fault. And employees may be penalised by having their disability benefit cut and, eventually, by being dismissed.
56In 2003 the first cohort of those who reported sick under this new protocol fulfilled the mandatory waiting period for filing a disability benefit. The disability benefit inflow rate over 2003 was 0.9 per cent of the labour force. This rate was 30 per cent lower than that in 2002, and reached its lowest level ever since 1967. As Figure 2 shows, the decline in the inflow rate was stronger among women than among men.
57Apart from the newly introduced structured case management of long-term absentees, two other factors may have contributed to this steep decline in the DI inflow rate. First, we mentioned that 2003 was the first year that experience-rated DI premium rates reached their structural level. In other words, in 2003 they became fully “biting”. Second, the downturn of the business cycle may have decreased absenteeism and reduced the number of cases of long-term sickness.
58Following the recommendations of a National Advisory Commission on Disability published in May 2001, the Dutch government drafted a set of bills. These are under discussion in Parliament but not yet enacted. They contain a drastic overhaul of the almost-40-year-old Disability Insurance scheme:
- – People are only awarded a disability benefit if they can be considered permanently and severely disabled.
- – Partial disability is covered by a separate wage-subsidy programme covering those who lose more than 35 per cent of their earning capacity but do not qualify for the first scheme (for the severely and permanently disabled). Government proposes to cover 70 per cent of earnings loss for those who get back to work but receive a lower wage. Firms are legally obliged to take out insurance to fund these supplements.
- – Those who cannot go back to their old employer and are unable to find employment elsewhere are entitled to Unemployment Insurance benefit. After lapse of this benefit, which replaces 70 per cent of the wage, unemployed partially disabled persons are entitled to a benefit corresponding to the degree of disablement multiplied by 70 per cent of the minimum wage. If this small amount brings their household below the social minimum level, they are entitled to an additional welfare benefit calculated against their partners” income.
- – The second group (with a capacity loss of less than 35 per cent) is entitled to support from their employer to stay in employment. In case of unemployment they are treated as regular unemployed persons who eventually receive means-tested welfare.
- – Expectations are that these two schemes together will reduce the inflow rate by two-thirds. Only if the inflow rate into the first scheme declines significantly will the benefit for severely and permanently disabled increase to a level of 75 per cent of pre-tax earnings.
- – As mentioned in section 3, as of 2004 the mandatory waiting period for Disability Benefit application is extended from one to two years. This makes the sickness benefit payment period for employers one year longer but reduces the burden of experience rating correspondingly.
59The first lesson that can be learned from the Dutch disability experience is that social disability insurance is often used as a provision to accommodate social change. A good illustration is the increase in female incidence rates. The disability benefit scheme supported Dutch households in their transformation from the traditional single-breadwinner type to the modern dual earner type. Similar uses of disability benefit schemes can be seen in Eastern Europe, where they ease the transition to a market economy.
60The drawback of using disability benefits as a “soft” childcare or unemployment provision is that it hides the lack of targeted, more cost-effective provisions and postpones their introduction. Meanwhile, huge, unfunded financial liabilities are created given the lengthy average duration of disability benefit dependency. The Dutch case is a good illustration both of the size of such liabilities and of the political problems involved in changing an entitlement-oriented disability policy.
61Nevertheless, in the face of an ageing workforce the Dutch government took a series of drastic steps from 1993 onwards. The second lesson is that many of these steps proved successful. In particular, the requirement that sickness and disability risks be borne where they can best be influenced – i.e. at the employer level – proved to be a fruitful approach. But the fact that management of absenteeism was strongly helped by the recently introduced legally binding protocol emphasising work resumption and prevention of long-term disablement shows that private provision of social insurance also requires rules and regulations in order to balance market efficiency and the social goals of disability policy.
A third lesson to be drawn from the Dutch experience is that the special features of the disability benefit scheme also proved to be its weakness. In 1967, the Dutch chose to integrate the general disability benefit scheme and the work injury scheme. They took the most generous of the two – the work injury scheme – as the model for a social insurance programme that covered all disability risks, independent of their cause. The framework of seven disability categories was taken from the work injury scheme. Applying this framework to all kinds of disability contingencies, including an increasing number of diffuse complaints, made the system weak and uncontrollable. The current proposal to cover only those that have hardly any productive capacity left is a logical last step in curing the Dutch disease.
Overview of acts and regulations relating to disability insurance and sick pay, and results and events
Philip R. de Jong: University of Amsterdam and Aarts, De Jong, Wilms and Goudriaan
Public Economics bv (APE), the Hague, Netherlands. Edwin L. de Vos: TNO, Hoofddorp, Netherlands.
This statistical definition is used by the national bureau for statistics (CBS, CBS-Arbeidsgehandicaptenmonitor 2003, to be published in 2005). In section 5.5 we present another definition of Arbeidsgehandicapten, as the target group of the Act on Reintegration of Work Handicapped Persons (REA) introduced in 1998. This population of work-handicapped persons is smaller by 250,000 than the self-reported vocationally handicapped population described in this section.
A separate programme covering the self-employed was abolished as of July 2004.
In 2005 the pre-tax minimum wage is €16,442,24 per year.
The duration of entitlement to wage-related partial unemployment (insurance) benefit is limited. It is based on the length of a person’s work record, with a maximum of five years. After lapse of this benefit a social assistance benefit may supplement the partial disability benefit up to the social minimum income to which the beneficiary’s household is entitled. Entitlement to this benefit is calculated against household means.
Dutch early retirement programs have no statutory basis; they emerged as an element of collective-bargaining agreements between trade unions and employers in 1975. The tremendous growth of early retirement plans since, the expected fiscal pressure of an ageing workforce, and benefits being paid out of pay-as-you-go funds, have called for changes in these actuarially unbalanced programs. An increasing number of these – collectively-bargained – plans are now being transformed into capital-funded flexible pension schemes with a much closer link between contributions and pension rights. These changes are likely to boost interest in the disability benefit option.
Transforming Disability into Ability, Paris, OECD, 2003.
“Broad” includes disability benefits, sick pay and work-injury benefits.
Disability beneficiaries are measured in full-benefit equivalents (i.e. corrected for partial benefits).
In 1976 the disability scheme was broadened. From then on it also included those handicapped in youth and the self-employed. The absolute numbers quoted are not corrected for partial benefits.
Figure 1 is based on data from the Social Insurance Institute (Kroniek der sociale verzekeringen, UWV) and from the National Bureau of Statistics (CBS, Statline).
Philip de Jong and Vincent Thio, “Donner versus Veldkamp’, APE report 53, October 2002 (in Dutch).
OECD Employment Outlook, Paris, 2002, p.69.
As of July 2005 the obligation to contract an occupational health service agency will terminate. Thereafter, the company, or its sick pay insurer, will decide what services it needs to prevent and manage absenteeism.
T.J. Veerman and J.J.M. Besseling, Prikkels en privatisering, The Hague: EBI, 2001, p.60.
T.J. Veerman, E.I.L. M Schellekens, J.F.L.M.M. Dagevos, J.A. Duvekot, F. Marcelissen. P.G.M. Molenaar-Cox, Werkgevers over ziekteverzuim, Arbo en reïntegratie, The Hague, EBI, 2001, pp.22-27.
Ph. R. de Jong and M. Lindeboom, “Privatisation of Sickness Insurance: Evidence from the Netherlands’, Swedish Economic Policy Review 11 (2004), pp.11-33.
Pierre Koning, “Estimating the impact of experience rating on the inflow into disability insurance in the Netherlands”, in CPB discussion paper no. 37, The Hague, CPB, Netherlands Bureau for Economic Policy Analysis, August 2004.
Clearly, the group that we described in section 2 is larger, and based on self-reports through a national survey. The difference is accounted for by the fact that the survey data in section 2 contain the total working-age population, including the non-employed without any work history, workers in sheltered workshops, recently disabled self-employed persons, and former recipients whose disability benefit ended more than five years before (E. L. de Vos and C. Smitkam, “Routekaart naar subsidies en sancties bij reïntegratie’, deel 3, Stecr Platform Reíntegratie, 2005)…
G.J.M. Jehœl-Gijsbers and C.G.L. van Deursen, Reïntegratie bij arbeidsongeschiktheid, Amsterdam: UWV (Social Insurance Institute), 2003.
A comparison among ten European welfare states, the United States and Japan, shows that the Dutch system of job protection during sickness is (still) much stronger than in any of the 12 countries (H. Bakkum and S. Desczka, De Nederlandse WAO in internationaal perspectief, The Hague, Ministry of Social Affairs and Employment, Werkdocument no. 241, 2002, pp.15-16).