Cash or in-kind transfers against financial problems?

Since 1996, every person in Switzerland has had to take out health insurance. At the same time, health insurance premiums are not dependent on income, which represents a major financial burden for low-income households. In order to ensure solidarity between people with different incomes, the cantons must support insured persons in modest financial circumstances with premium reductions. Today, the cantons and the federal government spend around five billion Swiss francs every year on social benefits, and more than one in four insured persons have their premiums at least partially paid by the state.

However, as with any means-tested support programme, the question arises as to how these transfer payments reach the recipients. Some cantons have decided to transfer the amount directly to the bank account of the respective recipient at the beginning of the year. With these so-called "cash transfers", recipients of premium reductions then pay the full premiums themselves, just like all other insured persons. Other cantons, on the other hand, transfer the money directly to the health insurance companies, which reduce the recipients' premium bills accordingly. The aim of these so-called "transfers in kind" is to ensure that the support is actually used for health insurance premiums. So far, however, there has been little scientific evidence worldwide as to whether cash or in-kind transfers are more effective in preventing financial problems.

Nevertheless, the Swiss parliament decided in 2012 to only allow premium reductions as transfers in kind. This meant that all cantons that had previously used cash transfers had to switch their premium reductions to transfers in kind from 1 January 2014. However, nothing changed for the other cantons, so they can act as a control group, as in a clinical experiment. From a research perspective, this is a stroke of luck, as such a "natural experiment" makes it possible to analyse the causal effect of in-kind and cash transfers on financial problems. By comparing payment deficits in the reform and control cantons before and after the reform, we can separate the direct effect of the payment method from general cantonal differences. This would not have been possible with the simple cantonal comparison mentioned above. With such simple comparisons, there is therefore a risk of drawing the wrong conclusions.

Measuring payment difficulties with premium invoices before and after the reform is not a problem in terms of data technology. However, it was much more difficult to differentiate between those receiving premium reductions and those not receiving them. One of the main reasons for the global lack of empirical evidence regarding the effect of in-kind and cash transfers is that, in the case of cash transfers, two data sources are required that are generally not linked. On the one hand, such an analysis requires data on payment difficulties, which normally occur in private companies. However, information on which persons receive support payments and to what extent is held by the state authorities - at least in the case of money transfers. We also encountered this obstacle in the case of premium reductions in the cantons with cash transfers until 2013. As the recipients there received unreduced premium invoices, the authorities did not inform the health insurance companies whether an insured person was receiving a premium reduction. Unfortunately, linking administrative data with other data sources is hardly possible in Switzerland - even for research purposes. As a way out of this data problem, we decided to estimate the probability of receiving premium reductions for each customer in all cantons in each year. Using a combination of several machine learning algorithms, we were able to predict who would receive premium reductions.

We were now able to take advantage of the harmonisation of the payment system in 2014 to compare the development of financial problems among premium reduction recipients in the control cantons with those in the reform cantons before and after the reform. For this purpose, we analysed a total of 22 million premium invoices from over 600,000 insured persons between 2012 and 2019 to determine whether a reminder or even debt collection was initiated. As can be seen from the two charts, the reform significantly reduced the probability of reminders and debt collection. The transfer to the health insurance fund has reduced the probability of reminders by 20% compared to money transfers, while the probability of debt collection has been reduced by over 12%. These improvements not only benefit the households concerned, but also reduce the burden on taxpayers, as the cantons ultimately have to pay 85% of unpaid premiums.

We were able to show that in-kind transfers are much better at preventing financial problems with subsidised goods. But could it be that the difficulties have simply shifted to other household expenditure? For this reason, we also analysed the invoices for cost sharing for medical services. For these rather unexpected (and never subsidised) expenses, we found no change in payment behaviour before and after the reform. We can therefore say with great certainty that transfers in kind lead to healthier household finances overall than cash transfers. It is therefore important in the political process not only to discuss the amount of state support payments, but also how they are organised. Changing the payment method could prevent a lot of suffering for recipients without costing an additional franc of taxpayers' money. Even if Parliament was right in this case in 2012 despite the lack of evidence, this should not obscure the fact that without the promotion of research and the consistent incorporation of the findings into healthcare policy, major welfare gains for all could remain unrealised.