I have often used the 'belling the cat' metaphor as shorthand for the difference between high concept strategy (always extremely easy) and actual implementation (always extremely difficult). I'm way overdue for a concrete example. Given the current discussion about health care reform, the 'notch effect' is a good illustration.
The notch effect bedevils social welfare schemes. I first learned the phrase in a book published by Daniel Patrick Moynihan in 1971, 'The Politics of A Guaranteed National Income'. It may be hard to believe post-Watergate, but the first Nixon Administration actually produced some innovative ideas. Many environmentalists would be astonished to learn that the EPA was founded then. Moynihan, the resident token Democrat intellectual in the Cabinet, attempted a massive overhaul of the welfare system, based on the notion of a guaranteed base income, which was borrowed from Milton Friedman's concept of a negative income tax. The reform failed completely, under merciless fire from both Left and Right. Moynihan wrote about the experience in the book I mentioned. Good reading, although the ordinary arithmetic (sic - not math) becomes mind bending within the first hundred pages. In any case, it was in that reading I became aware of the Notch Effect.
Suppose you are a conscientious Out. You make sincere speeches to the effect that a nation that can create the Internet can provide decent education, housing, health care to all the uninsured Poor. Now, suddenly, you are an In, and you have to translate that goal into practice. How do you do so?
I'll use 1970 type numbers for convenience. Who exactly are the Poor? You are a reformer, not a mad man, so you eliminate able bodied adults who can support themselves even if they choose not to. You restrict the definition to the disabled, and families with dependent children, with income of less than $5,000, You structure a system of education, housing, and health benefits to remedy the effects of this poverty. Well and good.
It is here that the notch effect shows up. When you set the poverty line at $5000 and provide substantial benefits for those below, you create a notch at that level. Someone who earns $4,999 a year receives benefits worth a several thousand dollars. However, if that someone works one more hour at the Burger King, and end up with $5,001, he or she gets nothing. A 2 dollar difference makes a (say) $5,000 difference. You unutterable fool. You are going to pin down the poor to permanent welfare dependency, unless the family (improbably) can leapfrog to an income position at which it can do without the subsidized benefits.
That obviously won't do, so you are going to have to include some graduated scale according to which benefits are phased in, and then fazed out. Well and good, but your simple little system just went up several levels in complexity. (That's why the arithmetic in Moynihan's book began simple but became mind numbing).
And of course, we are only thinking of income as a simple, one dimensional variable. What about illegal and non cash income? Drug dealers? Madams? Applicants with huge undisclosed cash resources? That clearly won't do. You obviously can't effectively police this sort of fraud at the administrative level, but you can regulate against it and criminalize the receipt of benefits when such conduct is turned up by other police investigations? But that's a whole new set of regulations, and another layer of complexity.
Are you talking about one income or two? The core issue of American social aid systems. Are you going to determine poverty on the basis of individual or family income? Individual? You fiscal idiot. By that measure, every minor child qualifies, no matter that his daddy's rich and his ma is good looking. You'll break the bank.
Ok. so eliminate minor kids and determine their eligibility by reference to parent's income. Sorry, but you still have a problem. Every stay-at-home mom with a conniving soul has suddenly become eligible, no matter that her loyal husband is a rock star. You fiscally irresponsible rad-lib.
Ok, so let's include the husband's income. WHAT????? You've just disincentivized and demoralized the most marginal fathers in society, if the benefits you are offering equal or exceed the income the father can provide.
Ooooooooooookay . . . why don't we include a 'presumptive minim income' from the father, regardless of actual earnings, so that he has no motive to leave? WHAT???? You heartless bastard. You are going to deny or limit benefits to needy families on the basis of PRESUMPTIVE income? Money that is non existent? Don't you know how difficult it is for a single mother of limited education actually enforce her rights?
And on and on and on. What about parental support? What about families headed by single women with children of multiple fathers? And I am simplifying by treating housing and health as unitary variables. Easy enough with the basics, but what about the grey area? Take acne with adolescents. Maybe it's a single zit below the lip of an impossibly vain cheerleader, or maybe it's the kind of major disfigurement that turns a formerly well-adjusted child into a social misfit. How do you decide coverage? By pimple count?
I have made (I hope) my point, without (I hope) insulting anyone's intelligence. The target here isn't anyone on the Left or Right. The target is empty sloganeering. For what it's worth, I think there's a moral imperative to assist the poor, and in such a way - to the extent practicable - that they are not disempowered. (This is my version of belling the cat.) I think most Americans have exactly the same value. Persons who are skeptical about this program or that do not necessarily have different values. They may have a different view of just what the task of belling involves, and how well it will really work.
The notch effect bedevils social welfare schemes. I first learned the phrase in a book published by Daniel Patrick Moynihan in 1971, 'The Politics of A Guaranteed National Income'. It may be hard to believe post-Watergate, but the first Nixon Administration actually produced some innovative ideas. Many environmentalists would be astonished to learn that the EPA was founded then. Moynihan, the resident token Democrat intellectual in the Cabinet, attempted a massive overhaul of the welfare system, based on the notion of a guaranteed base income, which was borrowed from Milton Friedman's concept of a negative income tax. The reform failed completely, under merciless fire from both Left and Right. Moynihan wrote about the experience in the book I mentioned. Good reading, although the ordinary arithmetic (sic - not math) becomes mind bending within the first hundred pages. In any case, it was in that reading I became aware of the Notch Effect.
Suppose you are a conscientious Out. You make sincere speeches to the effect that a nation that can create the Internet can provide decent education, housing, health care to all the uninsured Poor. Now, suddenly, you are an In, and you have to translate that goal into practice. How do you do so?
I'll use 1970 type numbers for convenience. Who exactly are the Poor? You are a reformer, not a mad man, so you eliminate able bodied adults who can support themselves even if they choose not to. You restrict the definition to the disabled, and families with dependent children, with income of less than $5,000, You structure a system of education, housing, and health benefits to remedy the effects of this poverty. Well and good.
It is here that the notch effect shows up. When you set the poverty line at $5000 and provide substantial benefits for those below, you create a notch at that level. Someone who earns $4,999 a year receives benefits worth a several thousand dollars. However, if that someone works one more hour at the Burger King, and end up with $5,001, he or she gets nothing. A 2 dollar difference makes a (say) $5,000 difference. You unutterable fool. You are going to pin down the poor to permanent welfare dependency, unless the family (improbably) can leapfrog to an income position at which it can do without the subsidized benefits.
That obviously won't do, so you are going to have to include some graduated scale according to which benefits are phased in, and then fazed out. Well and good, but your simple little system just went up several levels in complexity. (That's why the arithmetic in Moynihan's book began simple but became mind numbing).
And of course, we are only thinking of income as a simple, one dimensional variable. What about illegal and non cash income? Drug dealers? Madams? Applicants with huge undisclosed cash resources? That clearly won't do. You obviously can't effectively police this sort of fraud at the administrative level, but you can regulate against it and criminalize the receipt of benefits when such conduct is turned up by other police investigations? But that's a whole new set of regulations, and another layer of complexity.
Are you talking about one income or two? The core issue of American social aid systems. Are you going to determine poverty on the basis of individual or family income? Individual? You fiscal idiot. By that measure, every minor child qualifies, no matter that his daddy's rich and his ma is good looking. You'll break the bank.
Ok. so eliminate minor kids and determine their eligibility by reference to parent's income. Sorry, but you still have a problem. Every stay-at-home mom with a conniving soul has suddenly become eligible, no matter that her loyal husband is a rock star. You fiscally irresponsible rad-lib.
Ok, so let's include the husband's income. WHAT????? You've just disincentivized and demoralized the most marginal fathers in society, if the benefits you are offering equal or exceed the income the father can provide.
Ooooooooooookay . . . why don't we include a 'presumptive minim income' from the father, regardless of actual earnings, so that he has no motive to leave? WHAT???? You heartless bastard. You are going to deny or limit benefits to needy families on the basis of PRESUMPTIVE income? Money that is non existent? Don't you know how difficult it is for a single mother of limited education actually enforce her rights?
And on and on and on. What about parental support? What about families headed by single women with children of multiple fathers? And I am simplifying by treating housing and health as unitary variables. Easy enough with the basics, but what about the grey area? Take acne with adolescents. Maybe it's a single zit below the lip of an impossibly vain cheerleader, or maybe it's the kind of major disfigurement that turns a formerly well-adjusted child into a social misfit. How do you decide coverage? By pimple count?
I have made (I hope) my point, without (I hope) insulting anyone's intelligence. The target here isn't anyone on the Left or Right. The target is empty sloganeering. For what it's worth, I think there's a moral imperative to assist the poor, and in such a way - to the extent practicable - that they are not disempowered. (This is my version of belling the cat.) I think most Americans have exactly the same value. Persons who are skeptical about this program or that do not necessarily have different values. They may have a different view of just what the task of belling involves, and how well it will really work.
The Notch Effect illustrates the complexity of making policy, because people's lives are complex. In a way, it speaks to the need for technocrats - not that technocrats should make or choose policy, but (if you're trying to craft policy on a good faith basis) you will need the help of people who can disaggregate complexity and show you unexpected benefits and unintended consequences.
My views of the Bush 43 Administration were partly shaped by opposition to its values, but we live in a democracy and the other side wins more often than I'd wish. But it was also shaped by the knowledge that there were very few competent technocrats who could influence policy.
Those who were competent tended to be more ruthless bureaucratic fighters (David Addington, for example) than people like Greg Mankiw.
The Ur-Text for this is Ron Susskind's article in Esquire on the education of John DiIulio:
"There is no precedent in any modern White House for what is going on in this one: a complete lack of a policy apparatus. What you've got is everything -- and I mean everything -- being run by the political arm. It's the reign of the Mayberry Machiavellis."
Posted by: Andrew | September 30, 2009 at 04:04 PM