Wednesday, August 01, 2012

Friedman, freedom and paternalism

Milton Friedman turned 100 yesterday. Over on Skepticslawyers, Lorenzo's Friedman centenary commemorates the birthday with links to some other reviews. On Freedom and Flourishing, Winton Bates's What did Milton Friedman have to say about human flourishing? looks at Friedman's views on freedom. Another affectionate personal blogging view is provided in Remembering Milton Friedman on His 100th Birthday.

Professor Friedman exercised a considerable personal influence on my thinking when I was working in the Commonwealth Treasury, although my focus then was on his thinking about macro economics. 

One of the issues arising in comments was the question of freedom and paternalism.

David Drummond, the main writer articulating what I call New England populism, made the point in 1926 that all laws affected human freedom. He also warned against what he called the oppression of the minority. If Governments had powers, they would  use them. In democratic systems, this inevitably meant that minority groups were likely to be oppressed by the majority. Drummond was not opposed to Government action per se, far from it. His focus lay in the development of constitutional systems that would better match government to needs, that would control what he saw as the evils of power and oppression.

When I read Drummond's views in the early 1980s as part of my research into his life, they chimed with my own experiences. Among other things, governments and their officials believe that they are right, believe that the will of government should be observed and are inclined to exercise coercive power to achieve their ends even if that power is in fact being misapplied.

Thinking about the comments on Winton's piece in the light of my own views, I decided that I wanted to make a simple point in this post.

The provision of assistance to a group such as the unemployed or single parents on social grounds does not breech freedom except in the very broad sense that those payments limit someone else's freedom to spend their own money. Nor is that provision paternalistic because the state is not acting as parent, is not telling people what to do. Peoples' freedom is unaffected except to the extent that those receiving the assistance have choices that they would not otherwise have.

Things change, however, when behavioural rules are attached to the support. Requiring unemployed people to look for work is one class. Since the purpose of unemployment assistance is to provide people with a safety net until they find a new job, it is perfectly appropriate to require them to look. Whether the exact rules applied are reasonable is a very different question.

Welfare quarantining is a very different issue. Here we move into a new field, for we we have rules applied (we will dictate how you spend your money) that actually have little to do with the original purpose of the assistance. It may be a good thing that people send their children to school. We have laws and processes to bring this about. But this is a different issue from dictating how people should spend their money.

As a matter of general principle, the use of a public policy instrument designed to achieve one thing to achieve a different thing is always suspect. More broadly, when that use is intended to constrain behaviour, to make people conform to socially acceptable norms, then the smell of paternalism is in the air.


Evan said...

What would Friedman make of the unemployed being taxed by Centrelink at double the rate that multi-national corporates are taxed by the ATO?

I don't hear any noise expressed by his followers about this.

Anonymous said...

Good post Jim. Winton has also expanded his quotation of Mr Friedman very usefully. Thank you.


Winton Bates said...

My impression is that Friedman's followers do tend to be concerned about high effective marginal tax rates (EMTRs). There are difficult tradeoffs between the basic rate of benefit, EMTRs and the extent to which beneficiaries are harassed to look for work.

I now agree with you that, in terms of common usage. 'paternalism' does involve telling people what to do, even though being 'paternalistic' doesn't necessarily involve restricting liberty.

It is useful to maintain a fairly narrow definition of paternalism in order to distinguish paternalism from egalitarianism.

I would like to raise a few thoughts regarding welfare quarantining.
First, would your view be the same if the purpose of the measure is to ensure that children have adequate food, clothing and other necessities of life? I haven't checked whether any welfare payment does actually have that objective, but I would have thought that might be an important objective of some measures.
Second, is your view capable of being influenced by experimental evidence that quarantining works in terms of specified objectives. I don't have such evidence, I am just wondering if it is an 'in principle' objection.
Third, would you be influenced by evidence that taxpayers were more supportive of welfare payments in the presence of quarantining and willing to support higher levels of payment? Again, I don't actually have such evidence.

Jim Belshaw said...

Hi Evan. I broadly agree with Winton's response to your point, but would add the following.

Friedman was concerned with efficiency. His negative income tax proposal - - was designed to overcome the problem you talk about. And it is a real problem that is recognised but left aside as just too hard. And it shouldn't be!

kvd, re Winton's expansion, dialogue is good. We all learn.

Winton, questions. How very socratic! But questions are good.

You asked: would my view be the same if the purpose of the measure is to ensure that children have adequate food, clothing and other necessities of life?

The purpose of the family tax benefits is to assist in rearing children.If that is the intent and if the money is being spent on other things, then you can mount a case for a rule change so long as the rules address the intent.

You asked:is your view capable of being influenced by experimental evidence that quarantining works in terms of specified objectives?

Yes, but it depends on the specified objectives, on the nature of the inevitable trade-offs involved, and on the alternatives.

You asked: would you be influenced by evidence that taxpayers were more supportive of welfare payments in the presence of quarantining and willing to support higher levels of payment?

Probably not.

Anonymous said...

Jim I thought Winton's queries as to 'welfare quarantining' weren't directed towards the family tax benefit; but more towards the practical aspects of what used to be called the Northern Territory Intervention?

I guess I'll stand corrected, but on that basis your answers are a little askew, unfortunately. I'm only pressing this because I find it quite hard to reconcile my overall opinion with the particular that I assumed Winton was raising.


Winton Bates said...

kvd: In had in mind Noel Pearson's proposals. But I couldn't recall what they were specifically.

Jim's answers seemed to address the issues I had in mind - although I didn't particularly like his answer to the third question.

I think we should be sensitive to the possibility of taxpayers becoming more reluctant to provide welfare funding if the well-being of the recipients is not improved.

Jim Belshaw said...

The NT intervention is interesting because it generated three different models.

The first was the Noel Pearson model. As I understand this, this involved achieving community concensus on welfare quarantining. The Commonwealth then played a facilitating role. If my understanding is correct, that cannot be classified as paternalism.

The second was the initial NT intervention where, like the Cape York case, you had apparent significant adverse outcomes from spend on welfare payments. In this case, the particular intervention was intended to address a specific problem in particular communities but was imposed. Further, some elements in it were designed to force behavioural modification not originally connected to the payments themselves where other mechanisms were meant to exist but appparently failing - school truancy, for example. Now we have entered paternalism.

The third case involves the extension of quarantining via the pilots in other areas to a national approach. There are two elements to this. One is voluntary participation. Here the Commonwealth is creating a facilitating vehicle. It is hard to argue that this is paternalistic. However, the approach can be also be applied on a family by family basis where officials deem appropriate.

Now in both application and language, the paternalist element is quite clear cut. Thou shalt or else do the things that we regard as socially acceptable. This is nationally imposed behavioural modification supported in some cases by some reamrakbly moralistic langauge.

Winton, I will comment further on your third question tonight.

Jim Belshaw said...

Winton, just to amplify my response to your third question a little.

While you carefully qualified your question with the disclaimer that you had no evidence to support the question, it was a hypothetical one intended to clarify views, my position was influenced to some degree because there evidence that the opposite holds true. Specifically, people will support popular restrictions, but this doesn't translate in any obvious way to increase in support for the measure itself.

More importantly, if you look back across the last couple of hundred posts, you will see that I have been consistently campaigning against the tendency to adopt apparently low cost social controls that are popular where there is no evidence that they will have the desired effect. This affects my view.