Long Term Ambition Rivers
The Ministry of Infrastructure and Water Management has asked researchers from TU Delft, VU Amsterdam and ITS Leeds to use a Participative Value Evaluation (PVE) to measure the preferences of the Dutch about (effects of) river widening measures and dyke improvements in the context of the Long Term Ambition Rivers.
In the PVE, respondents had to choose between two variants for four locations along the Waal (Werkendam, Sleeuwijk, Gendtse Waard and Oosterhout) to meet the flood risk management standards: 'dyke reinforcement' or a 'combination between dyke reinforcement and river widening'. The first option is cheaper, but the second option also contributes to the development of nature and recreation. The participants were able to allocate a budget of 700 million euros. The participants had to choose between one of the two options at the four locations. If they had spare budget, they could spend it on other projects (road projects, projects that prevent water damage or projects that provide extra protection against flooding). After respondents had made their selection of projects, they were asked to justify their choice for each selected project. A demo version of the PVE can be found at: www.participatie-begroting.nl
Fixed budget PVE and flexible budget PVE
In this research we perform both a 'fixed budget PVE' and a 'flexible budget PVE'. A 'fixed budget PVE' means that individuals make choices with a fixed budget from the government (in this case 700 million). In a 'flexible budget PVE' participants can spend a budget from the government, but they are also given the opportunity to advise the government to spend more (less) money than the budget, thereby increasing (lowering) the tax. In a 'flexible budget PVE' participants allocate a government budget, but they can also influence their private income with their choices.
Preference for combination project
A total of 2,900 Dutch people took part in the PVEs. Some of these respondents (937) were specifically recruited in the Waal area. A first result of this study is that at all four locations where respondents had to choose between two variants to meet the flood risk management standards – 'dyke improvement' or a 'combination between dyke improvement and river widening' – the largest share of respondents for the combination of 'dyke improvement' and river widening'. Respondents who do and do not live in the Waal area make almost the same choices.
The choices of respondents were analyzed in a quantitative way using choice models. Based on these analyses, the best portfolio of projects – the portfolio that provides the most social benefit – can be determined. The model that statistically best determines the social costs and benefits of the investment options is an extension of the multiple discrete-continuous extreme value model (MDCEV). The model results show that the fact that the 'combination of dyke improvement and river widening' contributes to the development of biodiversity/nature is an important reason for the fact that respondents prefer this variant in order to achieve flood risk management standards over the variant 'only dyke improvement'. The positive effect on recreation also means that respondents prefer a combination variant
The best portfolio of projects is a portfolio in which the 'combination of dyke improvement and river widening' has been chosen at all four locations. Another result is that a major road project (A2 't Vonderen-Kerensheide) is not included in the top 10 portfolios. The results of a PVE can also be used to determine how high the additional costs of combination variants (or how low the influence on variation of biodiversity and recreation) should be to ensure that no more combination variant is included in the optimal portfolio. This can be investigated by performing sensitivity analyzes on the project effects on the basis of which the optimal portfolio has been selected. We performed four sensitivity analyses.
After respondents had made their selection of projects, they were asked to justify their choice for each selected project. The two experiments yielded approximately 10,000 motivations. These motivations are manually coded and classified into different categories. The qualitative results confirm that biodiversity and recreation are important factors in the choice of respondents. Another observation is that respondents' preference for the 'combination of dyke improvement and river widening' decreases when residents experience nuisance from the combination variant. The main reason for the fact that a relatively large proportion of respondents at the Oosterhout location opts for 'dyke improvement' can be traced back to the prevention of nuisance for households. In order to make the combination variant possible at this location, households might have to move. Several respondents explicitly stated that they find any mandatory relocation particularly problematic because it is not necessary. An alternative is available in which this nuisance does not occur (dyke reinforcement). Incidentally, it seems that citizens find it particularly problematic that there is nuisance for households. The number of households that will experience nuisance seems less important.
Usability of the results
The results of this research can be used to determine the balance between dyke improvement on the one hand and river widening in the long-term ambition rivers on the other. The social value of 'combinations of dyke improvement and river widening' is particularly greater than 'only dyke improvement', if the combinations have a positive influence on nature and recreation. This study shows that the Dutch are prepared to make sacrifices to ensure that flood risk management can be safeguarded with combination variants instead of dyke improvement variants. The results also provide insight into the specific projects that could be cut back to enable combination variants. For example, citizens are prepared to sacrifice the A2 't Vonderen road project to ensure that resources are released for combinations of dyke improvement and river widening. Furthermore, the results of this research can provide insight into the extent to which the Dutch support a tax increase to enable projects that are part of the PVE. In this PVE, the average chosen budget (688.36 million euros) was below the starting budget of 700 million. So there seems to be no support among the Dutch for a tax increase to enable projects that are part of the PVE on top of the 700 million euros of tax money that they can spend in the PVE. The results of a PVE can also be used to determine how high the additional costs of combination variants (or how low the influence on variation of biodiversity and recreation) should be to ensure that no more combination variant is included in the optimal portfolio. Finally, the qualitative results of the PVE provide insight into the aspects that citizens consider important about the projects. The qualitative results can reveal new considerations, effects and problems surrounding a project and therefore enable policy makers to consider these aspects in their choice and optimization of projects. Policy makers can not only use the qualitative information and the quantitative information that the PVE provides as decision information, but the information can also be used to substantiate choices. The use of qualitative motivations of citizens in the substantiation ensures that the substantiation fits well with arguments that citizens consider important.
Positive experiences from respondents
At the end of the experiment, we asked respondents to score the experiment on four criteria. The evaluations of respondents are very positive (you will find the scores in brackets)
- I am convinced of my choice (89% agree; 18% neutral; 3% disagree)
- I think this is a realistic experiment; (61% agree; 28% neutral; 11% disagree)
- I think it is a good thing that the government is trying to involve citizens in making choices between dyke improvement or a combination of river widening and dyke improvement; (82% agree; 12% neutral; 6% disagree)
- This experiment provides the government with relevant information when making choices between dyke improvement or a combination of river widening and dyke improvement. (65% agree; 26% neutral; 9% disagree).