Why I’m not holistic
A lot of New Zealand policy discussion and debate is holisitc – we say we want a set of outcomes, these outcomes are desirable, these are the outcomes we need to achieve. In essence, they are saying there are a set of states “we” could be in as a society and we should strive to be in the “one” they are talking about.
Note: There are elements of holism that are ESSENTIAL, but I see them as part of reductionist philosophy as well. For example, the idea that the state of the world influences payoffs beyond the actions of individuals is fair. Also the idea that we can’t compute everything is acceptable. My argument is against the idea that we start with a “target” (top down) rather than starting with “guiding principles” (bottom up) when designing policy.
What is an example, lets say “sustainability”. The idea of being sustainable, of having sustainability, sounds nice – it is an overaching state that some people want us to have. If they are forced to define sustainability they will say it has a whole bunch of characteristics – however, why these characteristics are good is never touched – just the fact that it takes us to this “state” of the world that they normatively believe is desirable.
However, I’ve never, not even as a child, found such descriptions convincing. One of these states must be reducable to the sum of its parts, there must be a set of causes that lead to that state – either that or it is an impossible pipe dream.
This is part of the reason I found economics so attractive when I was young – just think of the production possibility frontier.
The PPF tell us that there are a whole bunch of states of the world where you cannot make someone better off without making someone else worse off – these are our “pareto optimal” states. However, there are a multitude of states below the PPF line where outcomes can improve, and a whole bunch of states above the line which are unachieveable.
Given this, we need to know how states come about, so that we know where they fall in terms of the PPF line. We may envisage a state that we truely believe is superior to the one we are in now – but without a reductionist model of how that state comes about it may just be a pipe-dream, in may be impossible to reach.
Unless we can describe how a state comes to pass, how can we talk about it? How can it be policy relevant? And yet, most policy discourse seems focused on pipe-dreams, and involves a complete unwillingness of trying to work out what the trade-offs involved with anything are.
As a result, I do not believe anyone when they say “New Zealand needs a measurable goal” or “New Zealand needs a target”. “New Zealand” just needs the opportunity to do the best it can with the limited resources it has, in terms of the happiness of its citizens. We can only achieve such an aim by using fundamental economic principles to illustrate trade-offs, not by setting over-arching targets of “matching Australia” or “sustainability”.
But Matt, what about if we have imperfect information about the causes of outcomes – if we just know that an outcomes is “good” why not target it!
This is the most common argument against my point.
I don’t like it. For one, if information is imperfect it may be the case that the poor information is better than using no information at all.
But lets ignore that. If the information available regarding causes is so poor that we don’t think we can use a reductionist method – then how the hell are we supposed to come up with policy that improves outcomes? If we don’t understand causes, how can policy makers introduce policies that the “know” will ex-ante make people better off.
I do not believe that we can just pick an outcome, say that it is achievable, say that it is better than what we have now, and then make it happen – all in the absence of any information. People that do this (and there are many) are working off faith, and I’m sorry but I fear that these people will damage peoples lives with their good intentions – not help them.
Conclusion
I can often tell the difference between my non-economist friends and my economist friends by the degree of “holistic argument” they are willing to make. Both groups are intelligent, beautiful, people – but the structure of the discussions I have with them differ, partially due to training, and partially due to self-selection in the types of people who will be willing to do a subject like economics.
I have found that some of my non-economist friends get frustrated because I’m putting their discussion in “my framework” – it frustrates them, they feel that by doing this I’m losing part of their point. On my side I can’t understand why you wouldn’t use a framework that captures scarcity, trade-offs, incentives, and choice – and given that ANYTHING can be justified in an economic framework I do not feel like I’m missing anything. In fact, I believe that I am just making their assumptions transparent.
Now this is all just an explanation for why I debate and discuss the way I do – it doesn’t mean I’m right in any sense of the word, but my very nature of discussion and analysis is strongly reductionist. Note that I used words like “believe” and possibly even “should” all through this – it is because I recognise that my choice of method could be seen as, in itself, a subjective choice. As a result, I used terms that made it clear that I do this because I believe it – not because it is some mystical truth passed down from the gods.
This is a cool post, thanks for putting it up. I have two points. The first is related to your conclusion:
“I do not believe that we can just pick an outcome, say that it is achievable, say that it is better than what we have now, and then make it happen – all in the absence of any information.”
I think this is a caricature. Do the “intelligent, beautiful people” who make holistic arguments full of all those empirical claims really have no information at all to support them? Not in my experience.
Second, I want to ask you about the “guiding principles” that you say are the building blocks of a bottom-up policy framework. Do those guiding principles have any moral content, talking even implicitly about what is desirable, or are they only statements of empirical regularity?
If they have moral content beyond their empirical content, then they fall victim to many of the same critiques you make of targets;
If they are simply empirical regularities, then on their own those guiding principles cannot help a community pursue much other than nihilism. (Another way to think about this point: try making policy recommendations armed *only* with Mankiw’s 10 principles. I don’t think you will get very far – maybe the most you can say is “don’t print too much money” or “use markets most but not all the time”. Those policies raise more questions than they answer.)
Rightly, you argue that having a destination but no map is a ticket to a costly random walk. But your own proposals look to me like a guy with a map but no destination. In order to figure out how to make tomorrow better, first you have to define “better.”
@Rob Salmond
Hi Rob, thanks for your comment.
“I think this is a caricature. Do the “intelligent, beautiful people” who make holistic arguments full of all those empirical claims really have no information at all to support them? Not in my experience.”
The passage you are quoting is when I am talking in response to people saying to me “but what happens when we have poor information, how can we say we need to rely on causal mechanisms here when something is obviously good”. The intelligent, beautiful, people I describe are not the same people that throw this critique at me 😉
Now I agree that these people do have a reason, there is some form of causal mechanism there – but it is just my opinion that we should flesh out the entire argument before reaching our goals or targets, and I feel that not aiming to reduce things down is akin to ignoring this.
“Do those guiding principles have any moral content, talking even implicitly about what is desirable, or are they only statements of empirical regularity?”
Guiding principles in this case would be a methodological framework to work from – in its most raw form it might be value free, but of course without adding moral judgments we could never reach a conclusion from such a framework.
I would say the bottom up way of forming policy is first to come up with a framework that describes, and aim to add understanding of what happens. This is both a theoretical and empirical endevour – this is what economists do. Once that is available it is time to start chucking in moral judgments, adding views on value – and only then can we reach a conclusion, only then can we be told what is a “preferable” outcome. This is not what economist do – this is the realm of policy. However, it is important that the policy analysts first have the background provided by the economic method.
My preferences for such a method is that it is transparent – and it gives us some idea regarding “why” things happen. Furthermore, as I said in the post, it avoids the case where we decide to target something that is unachievable.
“Rightly, you argue that having a destination but no map is a ticket to a costly random walk. But your own proposals look to me like a guy with a map but no destination. In order to figure out how to make tomorrow better, first you have to define “better.””
But surely the goal of government is to maximise the social welfare function. I don’t see why we need a target outside of this – why do we need to aim to “catch Australia” without actually thinking of the costs and benefits associated with that “goal”.
Policy should (here I go being normative again) be based on incremental improvements based on information regarding trade-offs, the revealed preference of individuals, and the democratic mandate involved. In a world as uncertain and volatile as ours, I see no need to decide on a destination for the country – I just want sound policy based on strong economic fundamentals and the desire of the people, when those things are in place we will find ourselves on the edge of our production possibility frontier 😉
It sounds to me like we agree that you need both goals and an analytic framework in order to make policy. I apologise if my reading of your original post was wrong on that front.
For you, however, the overall goal is so obvious that it need hardly be debated: “But surely the goal of government is to maximise the social welfare function.”
I think that answer is only trivially true, and I say “trivially” because nowhere in that goal is there any statement of what goes into the social welfare function, nor about the timeframe over which we assess it. Stating a goal that way raises more questions than it answers.
In fact, you can think of those other target-setting exercises as revelatory of what is in our social welfare function. If we say collectively that we want to “catch Australia,” then you can infer that forms part of our social welfare function. The community gets to say what it wants. It could want anything, feasible or not. That is the nature of a welfare function – there are no theoretical legitimacy bounds on what a community can desire for itself.
And then policymakers, armed with their analytical frameworks and so on, are charged to try and deliver whatever-the-community-wants as best they can.
That sounds like a realistic version of a policymaking process to me, where the community at large has the agency to define its own welfare function, and then policymakers try to maximise it.
Trying to discourage people from setting themselves silly, unrealistic goals is an attempt to deny them the ability to define their own welfare function. I don’t think that is what you want to do, your subsequent paragraphs make that clear. But it is an unintended consequence of any attempt to proscribe the community in terms of what it should ask for.
@Rob Salmond
I definitely agree that people in society are allowed to intrinsically value a goal – my issue is when policy makers define goals without placing them within an appropriate reductionist framework.
I like the idea that communities will describe things, and provide information regarding things, that they value – areas where no sense of pricing is clear, or where implicit value is hard to define. And I would never argue against people illustrating preferences in this way.
My concern solely stems from policy actions that are based on aspirational goals – as setting such a goal and doing everything in your power to achieve it ignores:
1) The costs of achieving that goal
2) The actual benefit associated with said goal
3) The achievability of said goal.
My argument is in no way against people having preferences – it is just saying that I don’t believe in policy that is justified on the basis of goals.
“The community gets to say what it wants. It could want anything, feasible or not. That is the nature of a welfare function – there are no theoretical legitimacy bounds on what a community can desire for itself. ”
The is one side of the coin. There is no limit on what the individual desires either – but of course income limits what the individual can actually do. In the same case, there is no limit on what individuals and communities in society desire – but there are limits on the set of resources available to them.
As a result, government provides an institution that allows society to adjust the allocation of resources on the basis of external preferences – that is fine. But, the trade-offs associated with this need to be apparent. The social welfare function needs to be maximised subject to the constraint of what is achievable – and policy action should be both subject to, and transparent regarding, these inherent limitations.
“Trying to discourage people from setting themselves silly, unrealistic goals is an attempt to deny them the ability to define their own welfare function.”
Not at all. Discouraging policy makers from telling society that unrealistic goals are possible – and forcing them to be transparent about the trade-offs that will be faced – HELPS individuals make choices, and would surely help improve the determination of the social welfare function.
“But it is an unintended consequence of any attempt to proscribe the community in terms of what it should ask for.”
There was no prescription regarding the preference of individuals in this piece – there was a prescription regarding the promises of politicians and the behaviour of policy analysts. Aspirational goals are likely to obfuscate trade-offs – making it harder for society to reveal what it wants, and in turn placing additional constraints on the democratic process. A refusal to discuss the underlying trade-offs involved is relatively undemocratic – and this concerns me.