Reforming welfare states is no task for the faint hearted. It is a process fraught with political hurdles: dedicated, committed, entrenched interest groups often outmaneuver a larger, less committed public in pressuring the state to mold policy to suit their wishes. In many Western European countries, in particular, the task of trimming entitlements or reducing the size of the state—also called retrenchment—is arduous. As Paul Pierson writes in Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment:

Retrenchment is a distinctive and difficult political enterprise… Welfare states have created their own constituencies. If citizens dislike paying taxes, they nonetheless remain fiercely attached to public social provision… Voters’ tendency to react more strongly to losses than to equivalent gains also gives these programs strength. Retrenchment advocates thus have their work cut out for them. Almost always, retrenchment is an exercise in blame avoidance rather than in credit claiming.

The Dutch Disability Insurance system is a case in point. Recent reforms have finally reduced the number of people in the system, but only after decades of arduous political battles.

In the Netherlands, the labor disability system provides generous benefits to workers who are unable to work for reasons of physical or mental illness—that is, those deemed “labor unsuitable” after medical examinations. When the entitlement was created in 1967, the motive was to help those who truly could not work for reasons of illness: after a year of illness, an employee no longer had to work and was eligible to receive payments of up to 80% of his last-earned salary. (The clock for “a year of illness” starts the day the illness does.)

Dutch disability reform has been a politically difficult process, fraught with accusations of cold-heartedness on one side and fiscal irresponsibility in giving benefits to recipients who are not truly disabled on the other.

The expectation at the time was that 200,000 people would make use of the program, but this proved to be an underestimation: as is well-documented by the Dutch NOS in an in-depth report, by 1976, 500,000 people were in the system. By 1985, 764,000. By 1990, 900,000 people, at a time the working population was roughly 10 million. Highlighting the scope of the problem, in 1990, then-Prime Minister Ruud Lubbers of the center-right Christian-Democratic Party declared that “Holland is ill.” He later added that he would resign if the number of WAO-recipients were to surpass 1 million.

Employers were contributing to the problem: a Parliamentary Report released in 1993 by the Buurmeijer Commission found that many employers, interested in reducing labor costs, often “dumped” expensive older employees into the WAO system, replacing them with cheaper, younger workers and passing the bill for providing for older workers onto the state. The WAO entitlement program, meant to help the truly sick avoid destitution and penury, had become for many employers a de facto early retirement system.

As researchers from the Amsterdam School for Social Science Research put it in 2003: “The WAO was very attractive as an early exit road for older workers… The WAO operated and still operates [as of 2003] as a hidden unemployment scheme.” Moreover, not everyone in the labor disability system was, in fact, unable to work. A leading Dutch magazine, Elsevier, noted in 2001 that “The sad thing is that of most people deemed unfit for labor it’s barely possible to diagnose what exactly the physical ills are that make them unsuitable for labor. Besides the mental complaints [such as psychological fatigue and burnout] it’s often a case of vague back complaints…”

Elsevier also quoted former Labour Prime Minister Wim Kok, who several years ago said that “The Netherlands will always have a high percentage of persons unable to work, because we’re used to that.” Despite that example of complacency, successive governments have tried for decades to reform the entitlement. In 1985 the government reduced the maximum payments for the disabled from 80% of one’s last-earned income to 70%. In 1987, the government decided that people who were partly “disabled” and partly unemployed would henceforth receive unemployment benefits, not the more generous disability benefits. Additional reforms in 1991 mandated that people unsuitable for labor under the age of 50 were only eligible for a limited period of full disability payments: the reforms, later partially turned back, provoked a storm of protest. Elske ter Veld, the Deputy Minister of Social Affairs of the Labour Party, was furious and proclaimed: “Whoever wants to abolish the WAO has to abolish me first!”

For Social-Democrats, the labor disability insurance system was seen as a treasure of the Dutch welfare state—any attempts to change it became emotional issues. The Labor Party in the early 1990s was almost torn apart by grassroots anger over reforms to the entitlement.

In 1996, after a decrease in the number of people in the entitlement to 850,000, the numbers of those in the system begin rising again. That year, the government decided that employers should have to pay the first year of disability benefits. The measure was meant to dissuade employers from using the system as a dumping ground for expensive, older workers. In 1998, employers that saw many of their employees enter the WAO system were penalized by the government through higher contributions into the system.

By 1990, 900,000 Dutch were on disability, at a time the working population was roughly 10 million.

All of these reforms, however reasonable they may seem to outside observers, were hotly contested in Dutch politics and led to deep divisions. It has been a politically difficult process, fraught with accusations of cold-heartedness (the left accused the right) and fiscal irresponsibility in giving benefits to recipients who are not truly disabled (the right accused the left). Writing in 2003, researchers lamented that “the unsuccessful re-structuring of Dutch disability policy took several decades.”

The reforms ultimately took place because a reasonable consensus had formed—shared by a critical mass in the political elite and the public at large—that the status quo was neither sustainable nor justified.

In 2004, all WAO recipients under the age of 50 were re-evaluated according to stringent criteria to determine their suitability to be a part of the entitlement program. Benefits were lowered or ended outright in 52% of the evaluated cases.

In 2006, a new law came into effect: under the new rules, only those who are too sick ever to be able to work again, such as patients in an advanced stage of cancer, multiple sclerosis or those with heavy psychiatric problems, are declared “fully unsuitable for labor” and eligible for the most generous payments.

These reforms, at last, achieved their objective: the influx of workers declared to be fully unsuitable for labor has dropped by 96% under the new rules, to just 3,800 workers in the year since the new rules went into effect. By the end of this year, if projections hold, the number of people in the WAO-entitlement system will have fallen to 587,500, the lowest level in decades and a drop of 27% compared to 2002.

Many people who have not worked for a long time due to illnesses that are less straightforward than physical diseases (such as mental burnout) claim that they really cannot work, and a number of them have ended up receiving benefits under the unemployment insurance system rather than working. Their incomes have suffered accordingly and the accompanying human toll has sometimes been significant. Similarly, a number of people ordered back to work, because they were no longer qualified to receive benefits under more stringent criteria, have been let go by their employers, though the total number of such cases is hotly disputed. It will take more time to see if those ordered back into the labor market (the “partially work disabled”) are indeed working, or, instead, receiving unemployment benefits.

But the reforms are a success, if the goal was to secure the system’s fiscal foundations for those who are truly unable to work. That was the raison d’être of the entitlement at its inception, and thanks to the reforms, the fiscal future of the program is more secure—those truly unable to work for reasons of illness will receive payments to protect them from penury and destitution.

While there is still room for trimming, under the new rules, the program is set to eventually reach a size where the system is fair, responsible, and sustainable. The difficulties of reforming it, however, illustrate the challenges of reforming welfare states even if the need to reform is recognized by many societal actors, and should serve as a warning to those interested in reforming American entitlements: it is likely going to take decades, and while plenty of bullets will have to be bitten by policymakers, there will be no easy political solutions.

Jurgen Reinhoudt is a research assistant at the American Enterprise Institute.