In its current form, it might (not) be. Interested?
Diana C. Mutz’s 2008 article might just prove to be one of the most important articles I’ve come across (thanks to my supervisor). It builds a very compelling case against the “grand theory of Deliberative Democracy.” I don’t agree with everything she has to say though; so after a quick summary I try to highlight one of these contentious points.
Between theory and practice
In the paper, Mutz tries to bridge the gap between theoretical/normative and empirical approaches to deliberative democracy. Now since deliberative theory is highly normative, it’s no surprise that there is such a gap; but it’s further widened by unfortunate attempts to conceptualize and operationalize deliberation (or, for that matter, democracy). Grand theories of what deliberative democracy is are not good empirical[ly testable] theories: they come with an assortment of insufficiently clarified concepts; they often fail to specify the logical relationships between these concepts; and, what’s not unrelated, they are often unfalsifiable. In addition, they often fly in the face of compelling previous evidence.
There is one sense in which every theory of democracy is unfalsifiable. That sense relates to their normative dimension. Such theories describe what democracy should be like, and there is no empirical test can refute a certain value preference.
But, as Mutz claims, models of deliberative democracy tie their normativity to empirically observable benefits: that is, they claim that deliberative democracy should be embraced not because it promotes certain values, but because its practice leads to better decisions, better citizens, or better societies. If deliberative democracy doesn’t produce tangible benefits, it is then a flawed theory, one that towards the realization of which we should not strive for.
The problem is that it’s hard to test whether or not deliberative democracy (further on I’ll just write “dd”) produces these results, because, given the looseness and vague definition of deliberation, one can always claim that the desired effects were not produced because we don’t actually have deliberation running. To put it in another way:
Given the empirical consequences that theories of dd imply, evaluating dd requires two steps:
1.) deciding whether a given instance of discussion (or a given institutional setting etc.) constitutes deliberation, and if so, then
2.) deciding whether deliberation actually brings about desired results.
But if you define deliberation as that particular state of things which unquestionably does bring about the desired results, then the theory becomes unfalsifiable. In this case, you don’t hypothesize that deliberation will bring about good results, but assume it will; and that assumption will trump empirical observation. (I.e. if you find no good beneficial contribution, that’s not because deliberation doesn’t work, but because you set the experiment up wrong, or because what was going on wasn’t, actually, deliberation.)
So Mutz’s suggestion would be to first decide what elements of that loose and monstrous set that is called “deliberation” actually do produce beneficial results; and then, once we have that in our hands, we can go on establishing what might become some grand theory of dd.
Opening the black box
And the way of starting to do that, argues Mutz, would be to set up experiments, in which individual factors of deliberation are tested with empirically sound methods (e.g. including control groups), in an attempt to see whether or not they bring about desired results. Once we’ve done that, we could combine the results, and see how various factors interact. For instance- supposing that “civility” in debates is something that brings about a greater awareness in participants of oppositional arguments, and supposing that “offering rational arguments” increases the parties’ willingness to compromise, we can design a test that tries to find out how these 4 variables (civility, rational arguments // awareness of oppositional arguments, willingness to compromise) hang together.
Maybe we’ll find out that there’s no need for deliberation to produce the desired effects – which wouldn’t even be so surprising – since there’s a great body of compelling empirical evidence suggesting that deliberation has to work (but it doesn’t) in the confines of reality and of the fact that it will always remain a social situation.
Two ways of measuring success
As it probably shows, I’m totally impressed by Mutz’s article. However, I think there’s one aspect of dd which she downplays. I try to explain this now.
There are at least two ways in which you can measure the success (or failure) of a model of dd. On the one hand, you can look at the end results – the decisions that are made, the state of society, the public opinion or contentment of citizens or the like. That would be the output view – of course you’ll still need normative standards, and even with those the evaluation of decisions might not be so simple given that they always bear future and unforeseen consequences which might only be judged appropriately in hindsight; but that’s a problem outside the scope of this inquiry. The point is- Mutz argues that dd should be judged from this perspective only.
“The whole reason deliberative democracy is normatively desirable is because it is thought to produce tangible benefits for democratic citizens and societies.”
It’s here that I would disagree.
There is another way to measure the success or failure of a model of dd; and that’s through looking at the process itself, and seeing if it conforms to some normative ideal. Mutz does mention this point, but ends up dismissing it; and I don’t think she’s right in doing so. The way I see dd – and I’m convinced I’m not alone in this – is that, while it does produce empirically observable benefits, these are rather a set of side effects. The real reason dd is normatively desirable is because it is thought to establish a procedure that can guarantee the (highest degree of) legitimacy of democratic decisions. Or, more precisely: dd is thought to establish a procedure which guarantees (to the extent that this is possible) that the single best solution is going to be found to any given problem that enters the field of the political. (Of course you can also reject the whole possibility of there being any best, or even best-in-the-current-circumstances, or best-we-have-right-now decisions.)
Now it’s important to stress that the best decisions are the best because of their own merits – as Habermas would say, because they can be supported with the largest number of the strongest (and most generalizable) arguments. Could a dictatorship produce such “best decisions”? By all means. As far as the outcome is concerned, it really doesn’t matter if a given decisions was arrived at in a deliberative manner of not.
But deliberation is still important, because that’s the best guarantee we have over the quality of the decisions reached. It’s a built-in quality check. This is because deliberation is aimed at creating the largest possible pool of opinions and arguments, and at testing each proposed solution (or opinion) against each of the arguments. Participation in deliberation is inherently good because this is a trivial and efficient way to enlarge this pool of opinions and arguments.
Of course this is still an assumption: we assume here that a larger number of people involved in a debate produces a larger pool of opinions and arguments. But this is a trivial step to take. If we accept only this much, then we already have a valid normative grounding of deliberative democracy.
A wild counterargument appears: can it not be that too much contribution is bad? that a too big pool of opinion is going to be detrimental?
Sure, that can be – it seems even likely that such is the case; for it is only possible to debate about things for so long, and to consider only such a small number of alternatives and arguments. Too many contributions can clog the system. We need deliberation to have some kind of a filtering and “argument management” system which makes sure that the best arguments will, in fact be found.
But that doesn’t overrule the basic idea that more – well: – ideas, or more suggestions, is a good thing. Filtering is a technical detail; creating ideas – presumably from various life experiences – is not.
So… what now?
Accepting all this means that we have one empirically observable variable (the number of participants (and that of the opinions and arguments offered)), tied firmly to the normative base of deliberation, which indicates the extent to which a necessary (but by no means sufficient!) condition of dd is fulfilled.
And this is what I’ll try to stick to in my current undertaking, which is trying to establish what normative implications various ideal types of democracy have towards the Media.
I’m not entirely sure just yet how exactly I’ll operationalize this “the more, the merrier”-principle, but the basic idea is that the presence of a wide pool of opinion and arguments is a necessary condition for successful deliberation. If it is met, deliberation can still fail in a number of ways; but if it’s not met, then deliberation cannot succeed – not from the “input” or “process” view, even if decisions reached will be judged perfect from the output perspective.
Any thoughts on this?
Mutz, Diana C. (2008): “Is Deliberative Democracy a Falsifiable Theory?” in Annual Review of Political Science, 11: 521-538.