Humphreys & McLellan [1] advocate policies for more effective substance use disorder service systems that reward in-treatment outcomes, rather than long-term effects, and give patients purchasing powers in a treatment market.

They write modestly that they do not know to what extent their (stimulating) conclusions about best strategies to improve treatment outcomes are relevant outside the United States and the United Kingdom. It is true that a demoralized staff with little professional competence and low status is not a problem that has been raised in Nordic discussions. Other matters mentioned have, however, also motivated centrally initiated reforms in our countries, such as a lack of medical competence and under-utilization of evidence within the systems. A concern for less resourced groups of substance abusers has also been expressed. Sweden has put much effort into implementing evidence-based treatment methods, and supported treatment of so-called heavy abusers. Integration between substance abuse treatment and (mental) health care are part of reforms or reform plans in Finland, Norway and Sweden [2–5]. When these ideas have been accompanied by earmarked state funding they have influenced the local systems—but with largely unknown effects.

Money matters most for reforms. There is a constant conflict between universalistic ideals and distribution of limited resources. In the Nordic countries, substance abuse treatment is financed mainly through taxes (with a mix of providers). The responsibility to provide various kinds of treatment in accordance with the needs of the population has been highly decentralized to the municipalities. At the local level, and particularly since the 1990s, with growing attempts to downsize the public sector, a primary interest in most reforms has been to decrease the costs of the services. Substance abuse treatment resources compete for (municipal) tax money with expenses for services, for instance, for the elderly or for daycare. Substance abuse treatment belongs to the social and health-care services that have the least support among the voters and politicians, and that will be the first to experience cuts when cuts are called for.

Since the late 1980s, market models have been tried out in a variety of versions in the substance abuse treatment systems in all Nordic countries [2,4,6,7]. Market models change the systems swiftly, as they will directly allocate resources, and cost-efficiency within substance abuse treatment is assumed to increase through competition and consumer influence. Market models also seem to be a safe route to shrinkage of the public sector, which is accused of being inefficient. Consumer choice is finally ideologically in line with what is regarded today as normal for the empowered service consumers in the welfare systems. However, we lack evidence. Market models in good times can empower well-resourced substance use disorder clients with fewer chronic problems, but the application of market models in substance abuse treatment has been followed by many problems that could have been anticipated.

The Swedish history of substance use disorder treatment shows that, in general, clients' possibilities to influence what treatment they received were greatest during the heyday of the universalistic welfare state, until the late 1980s. Those were the days with a centrally and locally planned good mix of services for all groups of problem users, with earmarked state subsidies to local treatment systems and with few financial restrictions. This was before the markets.

From the consumer's viewpoint, in reality choice today is often limited. It is not only true that in most smaller communities there are no real alternatives, but in larger localities competition has also, in many places, implied a streamlining and uniformity of services. Sometimes there are local monopolies and sometimes a choice only between places with Minnesota model programmes or between doctors who prescribe more or less buprenorphine [8]. Market models further create new bureaucracies, with the need to install financial control of purchasers and control of providers. Markets are not necessarily cost-efficient, and especially not from a population viewpoint, as the market will not meet all needs. The markets favour large groups, those that have more resources or solid public support and those that have better possibilities to reach recovery, however it is defined. The markets can have unintended, cost-increasing effects. Many opioid-using clients, if given the choice, will choose a doctor who subscribes more, rather than less, buprenorphine. It has also been pointed out that markets with competition can break up the continuity of care between different providers in a publicly planned system [2,9].

The authors call for realistic evaluations to accompany all new initiatives. It is easy to support this conclusion. It should also be applied to market models. We need more evaluations. We need to share more effectively the experiences we have.

Service systems are dependent upon decisions about distribution. The market is not a good way to distribute services between clients with different levels of information, varying degrees of real choice and varying possibilities to correct their wrong choices. It cannot substitute the political decisions that will distribute services to all those who need them.

There is no ‘quick fix’. We will have to accept that treatment systems change all the time, with changing problems and needs, that they should be evaluated constantly and publicly, with strong user involvement, and with open discussions about distributions of resources and social values. The market is not a solution.

Declaration of interests