Imagine you have walked back home from your local town market on a jasmine-scented Saturday morning with a bagful of the season’s harvest. In Northern California in the summer, that bag will probably contain some heirloom tomatoes, hothouse cucumbers, red bell peppers, Meyer lemons, and mint sprigs. As you sit to rest your feet, your mouth starts to water in anticipation of how these provisions will taste. They are meant to entertain guests over supper later in the evening, but you simply cannot wait and decide to steal a sampling of small pieces of each item.
How should you go about tasting the produce you have just bought? Should you cut up small chunks, bits, and squeezes of everything into a salad and bite into each ingredient separately? Or should you blend everything together into a summer gazpacho? Taking separate bites of a salad allows you to taste the integrity of each ingredient, then, as the need or inspiration hits, draw an organic impression of everything together, eaten as a "meal." With a well-blended soup, the organic sense of the meal comes first, then – if some element of the meal tastes especially funky or flavorful – you might be drawn think about the distinct characteristics of each ingredient.
This choice – with some generous abuse of metaphoric license – represents two contrasting approaches to deliberation that the European Commission explored in 2007. In deliberation, soups and salads are alternate approaches to fairly and accurately representing the considered opinions of a general population of interest, and the ingredients are the constitutive sub-groups of that population of interest. Demands to fairly and accurately represent the viewpoints of distinct sub-groups in pluralist societies – whether during deliberative moments or during other modes of political expression – are constant. How to achieve this fair and accurate representation, however, is not altogether clear.
The European Commission confronted this dilemma squarely in the aftermath of the 2005 referenda defeats for the EU constitution in France and the Netherlands. Faced with an unexpectedly large (at least to EU elites) gap between elite and mass support for the European Union, the EC committed itself to better understanding this elite-mass gap and to making a dedicated effort to bridge this gap. To this end, the EC called for a period of reflection and launched Plan D for Democracy, Dialogue, and Debate in October 2005. The stated goal of Plan D was to stimulate national debates and engage citizens in a EU-wide discussion about the EU. How does one reckon public opinion in a political union comprised of 27 member states?
Plan D provided funding to six major projects. Two of them attempted to mount direct participatory events on a scale that involved some citizens from all member-states in deliberations over their expectations from the European Union. These two events were the Tomorrow’s Europe deliberative poll (TE) and the European Citizens’ Consultations (ECCs). The TE and ECC events are potentially interesting to this blog’s audience because they draw a sharp line between on two different approaches to deliberative moments in diverse polities. Both approaches derive their legitimacy to the extent that they achieve a “mini-public” that fairly and accurately represents what we might imagine to obtain if the hundreds of millions of voter-eligible EU members were to reflect upon and deliberate about some issue of common interest. That is, legitimacy comes from serving effectively as a deliberative proxy.
Whither soup and salad in this contrast between TE and ECC? The Tomorrow’s Europe deliberative poll, organized primarily by the think-tank Notre Europe, is the soup, a blended gazpacho of EU opinion. Deliberative polling, a technique developed by James Fishkin, attempts to assemble a group of deliberators randomly selected and representative of the polity at large across all major demographic categories. Applied to the EU as a polity, TE organizers claimed to have randomly sampled 3,550 individuals from the 27 EU member countries (recruited in parallel by a single French polling firm). Out of this pool, 362 individuals were chosen (more or less randomly) to participate in the TE event in Brussels, October 12-14, 2007. The remaining 3,188 were surveyed as a “non-participant” comparison group.
TE’s approach to deliberation is a blended gazpacho approach because the “ingredients” from all 27-member countries were whisked together into “EU opinion” over a three-day weekend in Brussels. The construction of “public opinion” out of the TE event (the taste of the meal, if you will) is a fusion of the weighted mix of opinions from all member countries. Some ingredients were in large volume (e.g., participants from Germany and France) and some in vanishingly small orts (e.g., participants from Cyprus, Estonia, Luxembourg, and Malta). The key question for deliberative polling, as a method for making soup, is whether the “secret ingredient” – Fishkin’s trademarked approach to providing participants with information packets ahead of the event and access to policy experts during the deliberations – changes the flavors of the blended ingredients or leaves them essentially unchanged, no better or worse for the wear of talking through an issue under idealized conditions.
The European Citizens’ Consultations, by contrast, is more of a salad. While TE organizers committed themselves to a single EU-wide deliberation among a group defined as a “representative microcosm of Europe,” the ECC chose to convene 27 member country-specific deliberative events over a four-week period in February and March, 2007. As with TE, ECC organizers had their own “secret ingredient” – or to continue with our metaphor, their homemade approach to dressing a salad – to execute their 27 more or less coterminous conversations about Europe. Each deliberative event shared a boilerplate protocol and the recruitment of participants was randomized within each country. Notably, each event was directed by a local partner with knowledge and experience specific to each member state context. One inevitable consequence of this approach was a discretionary range in how randomization of samples was achieved and in how the ECC’s boilerplate protocol was implemented.
Importantly, however, the ECC approach insisted on constructed 27 distinct mini-publics first. Thus each event recruited between 28 and 200 participants in each member country, for a total of 1,477 participants in all 27 events. While some components of this salad approach were small and others large, as with the TE approach, each ingredient (that is, each country-specific event), however, was tasted separately. In short, the construction of “EU opinion” in the case of these ECC events is wrought from the parallel construction of 27 independent publics.
In a recent paper, Pepper Culpepper, Elena Fagotto, Archon Fung, and I examined several facets of the empirical and normative trade-offs entailed in this contrast between the TE and ECC approach to deliberation. Our ability to draw any strong conclusions about the strengths and shortcomings of these approaches is limited by the availability of primary data on how public opinion was constructed in these two events and what impact that construction had for subsequent policy debates on EU enlargement, the environment, social welfare, immigration, pensions, and the like. It is clear from the two experiences, however, that organizing pan-European mini-publics is a feasible enterprise.
Beyond feasibility, there are several other points of note. As with any governance mechanism that relies on voluntary citizen participation, there is a descriptive difference between participants and non-participants. Participants in both TE and ECC events tended to be more highly educated, more professional, and more “Europhilic.” Thus, while exacting costs may help to mitigate the distortionary bias in who participates and who does not, it cannot eliminate them entirely. This constraint is especially telling for TE organizers, who went to extraordinary lengths to strive for a random representation of the EU, but fell shy of the mark in several respects.
More substantively, the TE event found that deliberative polling increased participants’ willingness to make personal sacrifices in support of broad social welfare goals (e.g., saving pension regimes). At the same time, deliberative polling also appears to have significantly dulled participants’ appetite for further EU enlargement (particularly for admitting Ukraine and Turkey). Thus in one domain, deliberation appears to expand the “circle of we”; in another domain it appears to shrink.
In the case of the ECC, the result of convening 27 separate country-specific deliberations appears more inclusionary. Participants emerged from their deliberations with a more favorable regard for EU institutions and EU membership. Participants also appeared more attached to democratic norms as an element of EU decision-making as a result of the deliberations. Whether this substantive difference between the TE and ECC approaches survives the test of time and closer scrutiny, however, remains to be seen. Ultimately, in Europe (as elsewhere), what we know about the prospects for deliberative democracy is dwarfed by what we have yet to discover.
Photo credit: Flickr user prettydaisies