Research
Market analysis
Reforms of both the social security system and the health service in the Netherlands have raised issues related to market forces, including the following:
- How is the market performing in terms of prices, quality, accessibility, efficiency and innovation?
- Are there any market failures?
- Are there any failures caused by government intervention?
- What form of government intervention is effective?
- Is government intervention distorting competition?
Examples of research:
845: The market for childcare services, 2004.
817: No care(s) tomorrow?
791: Market delineation for the proposed merger of Hilversum and Gooi-Noord hospitals.
Evaluation, effectiveness and cost-benefit studies
The Health Care and Social Security section conducts both qualitative and quantitative evaluation studies. The former is based upon interviews carried out within a clear conceptual framework, as well as upon literature research. The latter may be based upon either pre-existing data or information we collect ourselves. In many cases, both qualitative and quantitative methods are applied in answering a research query. The trick in an evaluation or effects study is to be able to find a causal relationship between changes observed and the policy being assessed.
We are specialists in using quantitative methods to measure the effects of policy; techniques like difference-in-difference, regression discontinuity, instrumental variables, propensity score matching and hazard models have no secrets for us. Going a step further than effect measurements are cost-effectiveness and cost-benefit analyses, in which the effects of measures taken are weighed against their cost. Naturally, these studies are also found in our toolkit.
Examples of evaluations and effectiveness studies:
844: The way back: from incapacity to work.
Examples of cost-benefit analyses:
858: Costs and benefits of EQUAL projects: accountability document.
738: Costs and benefits of extramuralisation.
677: Costs and benefits of employment services.
Budgeting and costing systems
Private companies have an incentive to produce goods and services efficiently, something which highly regulated organisations do not automatically share. The trick, then, is to design a costing system which gives those organisations the right stimuli to operate efficiently and effectively. We are experienced in developing and calculating the effects of costing and budgeting models. Questions these can help answer include…
- Should costing be based upon input, output or effect (no cure, no pay)?
- How is the available budget established and distributed?
- What is the effect of a personal budget or a voucher scheme?
886: Socioeconomic status in the risk equalisation model for health insurers: what are the possibilities?
890: Off the shelf and into work.
817: No care(s) tomorrow?
Explanations, forecasts and simulations
The Health Care and Social Security section has considerable experience in explaining and forecasting developments. In so doing, we conduct both qualitative and quantitative research, as well as developing explanatory and predictive models. These can be used to make forecasts, either to predict what will happen if current policy is maintained or to simulate the effects of policy variants or changed behaviour. The models may be based upon pre-existing data or upon information gathered specially for the study. When producing estimates, it is sometimes necessary to measure consumer preferences. SEO has the techniques to do that: the conjoint analysis.
Examples of research:
907: A well-earned pension?
862: Dutch lessons from the Swedish pension reform.
757: Financial incentives for early retirement.
Measuring preferences
In a "normal" market, consumer preferences are revealed by the workings of the pricing mechanism. In healthcare and social security, however, that mechanism is largely absent and so our picture of what people want is muddied. Yet consumer preferences are often crucial to perform an accurate market analysis. In order to measure whether the market is giving people what they want, we need to know what that is. And the same often applies if we are to explain and forecast behaviour.
To establish consumer preferences, we use data based upon actual behaviour – so-called "revealed preferences" – as well as our own questionnaires asking what people want in particular situations. We use conjoint analyses, which force people to weigh up various aspects of a situation or a product. This approach provides a better picture of true, considered preferences than methods that involve asking consumers directly what their preferences are. Amongst other things, we have used this method to forecast the nature and extent of the need for childcare services for teenagers (report no. 608) and delineate the market in the light of a proposed merger of two hospitals (report no. 791). In that case, people had to weigh up medical quality, waiting time and travel time so that we could assess whether patients are prepared to travel longer if they receive better quality of care as a result.