In defense of the "there are no win-win government policies" claim

In the green jobs debate, decision makers and green jobs advocates seem to be looking for win-win government policies. I've pointed out that there are no win-win policies. Any government policy creates losers in the form of those who bear the costs of the policy. In other words, "there is no such thing as a free lunch."

Several readers find me confused. I reply:

Thanks. But I'm not confused.

Here is the rest of my reply in the comments section:

"Win-win" is invoked in the context of public policy. The "win-win" headline suggests that there are no costs. Hence, it is important for an economist to recognize costs.

"Gains from trade" results from private transactions and is irrelevant in this discussion

A Pareto improvement is by definition a "win-win". Or, at least, "win-no loss." Yet, there are no Pareto improvements in public policy which is why benefit-cost analysis uses the potential Pareto improvement criterion.

So, I should back off my "there are no 'win-win' policies" a bit. How about this? "there are no 'win-win' policies without compensation of the losers"

 

???

To provide background for this "partial retreat" from my hard line stance, here are some quotes from the Boardman et al., Cost-Benefit Analysis textbook:

An allocation of goods is Pareto efficient if no alternative allocation can make at least one person better off without making anyone else worse off. (p. 26)

If a policy has positive net benefits, then it is possible to find a set of transfers, or side payments, that makes at least one person better off without making anyone else worse off. (p. 27)

The implementation of policies almost always requires the use of some inputs that could be used to produce other things of value. ... The opportunity cost of using an input to implement a policy is its value in its best alternative use. (p. 28)

As long as analysts value all impacts in terms of willingness-to-pay and all required inputs in terms of opportunity costs, then the sign of the net benefits indicates whether or not it would be possible to compensate those who bear costs sufficiently so that no one is made worse off. (p. 29)

One could imagine the following decision rule for CBA: Adopt only policies that are actually Pareto efficient. ... CBA utilizes an alternative decision result with somewhat less conceptual appeal, but much greater feasibility, than the actual Pareto efficiency rule. ... A policy should be adopted if and only iff those who will gain could fully compensate those who will lose and still be better off. (p. 29)

In other words, there are no "win-win" public policies. There are public policies where the benefits greatly outweigh the costs but ... hmmmm, the costs are still positive (even if they are small).

Pointing out positive costs is not equivalent to having a "sufficiently jaundiced view" or "invoking a very counter-productive argument that there will always be some obscure loser out there. What's the point?"

To an economist, whenever someone advocates a government policy that has no costs (i.e., "win-win") this raises a red flag and initiates a bloodhound-like zeal to identify exactly what the costs are. It may turn out that the benefits of the "win-win" policy greatly outweigh the costs. But, oftentimes, you may find that the costs outweigh the benefits and the "win-win" headline is being used to obscure who might be the losers.

For a non-environmental example, advocates of tariffs might say that the tax on international goods is a "win-win." It generates government revenue and protects domestic jobs. However, it obscures the large, but individually small, costs to consumers (i.e., higher prices).

 

Comments

[Scott]

Your analysis is true ONLY if you limit it to the measurable economic impacts. If you broaden the analysis to include health impacts, there being tens of thousands of excess deaths every year due to environmental impacts, then I think it likely that your analysis will not prove true.

[John Whitehead]

Scott,

The positive health impacts are included as benefits in a benefit-cost analysis. The existence of positive health impacts may offset costs but it doesn't make them go away. This is an important distinction. In other words, 10 - 2 is very different from 10 - 0.

[odograph]

I saw someone say "economics is politics, but with added mathematics" this last weekend. Was the quote original?

(FWIW, much of this might be an instance of a much more commonly seen human tendency. Split brain studies etc. are able to show that after we have an idea we are very adept at naming rational arguments for why we have it. But as those experiments show, the belief can come before the rational machinery is engaged. This becomes a caution to all of us who argue positions and should teach us some humility. Where did we start? Which was the chicken and which was the egg?)

[John Whitehead]

Odo,

Huh?

[odograph]

Do you think your essay above depends on a narrow definition of "win-win?"

I might think that my new (now old) refrigerator was a win-win, if it could pay for itself in a few years. My electric company might think that encouraging such refrigerators (perhaps just with a "nudge" in modern parlance) was a win-win as well.

How deeply do we need to look for negatives? Is it possible that an exhaustive search is more rationalization than reason?

[Tim Haab]

Odo,

There is only one accepted definition of 'win-win' that I know of: everyone wins. You seem to be implying that the definition is: everyone I think matters wins.

It's really simple: Costs are costs.

[John Whitehead]

Odo,

Since it takes a few years for your refrigerator to pay for itself, the higher upfront cost of your energy efficient refrigerator reflects the intertemporal tradeoff between consumption now and consumption in the future. Given a positive rate of time preference, people prefer consumption now and must be paid to forego the benefit (i.e., the foregone benefit of current consumption is the present cost).

This is basic consumer theory. I'm not making it up, looking deeply or conducting an exhaustive search.

I didn't make up "there is no 'win-win'" or the notion that "jobs aren't benefits" in benefit-cost analysis. These are standard economics.

I'm not out on a limb or guilty of a lack of humility. (I'm actually very humble.)

The rational machinery came before the belief.

Pointing out that costs are costs doesn't mean that you ignore the benefits. Economists consider both benefits and costs!

[odograph]

I would say that "everybody wins" is an extreme definition.

If for instance Iran had no more money for uranium centrifuges, most US observers would not find it a useful disproof of our "win-win."

That might be too exhaustive a search ;-)

[odograph]

(Probably to normal humans "win-win" means that the top few articles of concern are satisfied, and that you have to hunt for negatives (which are minor when you finally find them under the bed in a dusty old box))

[John Whitehead]

Odo,

I would say that "everybody wins" is an extreme definition.

So, when you say "win-win" ... you actually mean that someone loses?

If so, weird and misleading. When I say that benefits exceed costs (or vice versa) there is no confusion.

[odograph]

I was answering Tim above, before seeing your last (1:41PM) post, John.

I'd say it's about what you headline and why.

Do you have serious costs to defeat the "win-win" in the public mind, or are you under the bed?

(I think the defense to my win-win, that other people might choose otherwise, somehow misses the mark.)

[John Whitehead]

I doubt that the coal industry counts the costs in this "win-win" policy as minor.

[John Whitehead]

Odo,

Who has the right to focus on the "top few articles of concern"?

Also, the point is that there may not be enough costs to "defeat" a "win-win." Pointing out costs doesn't imply that the pointer outer thinks they are bigger than the benefits. I've previously defined what you call a "win-win" a situation where the benefits greatly outweigh the costs.

Every policy has costs that should, and can, not be ignored.

[odograph]

Who has the right to focus on the "top few articles of concern"?

If we couldn't do that we couldn't make breakfast in the morning. ;-)

The counter argument to a particular win-win, might be "wait! this is a serious concern ..." That would be reasonable.

But to say there are no win-wins because there is always some negative (no matter how remote) is madness(or sophistry).

[John Whitehead]

Odo,

I don't quite understand your breakfast analogy. When I want bacon, eggs, toast and raspberry jam, I dont' leave out the raspberry jam. Only an idiot would eat toast without raspberry jam.

A good benefit-cost analysis will attempt to qualify all benefits and costs and quantify the major ones.

Your "madness (sophistry)" quip only suggests to me your lack of seriousness and/or a simple rejection of the "economic way of thinking".

[Blake]

I agree that there is no win-win in terms of policy, but maybe in the technological perspective.

[Scott]

John,

I assume the avoided health care costs are accounted for, but how does one account for suffering or not being dead?

[odograph]

How long do you spend thinking about the organicness of your raspberry jam?

To get what I meant by the breakfast thing it might help to have some exposure to early AI and cognitive views of human decision making. Early AI researchers had a very cognitive self-image, and an idea that we could make an intelligence by tallying all the necessary and logical little rules for everything.

It turns out minds don't work that way. They work, without internal gridlock, because they can ignore all of the multitudinous and minor questions hanging off every task or question.

... and so I'm not really impressed by obscure losses dredged up to disprove win-win. At some point it becomes counterproductive (or pathological).

[odograph]

BTW, I'll remind you that I've said if you can defeat a particular win-win with a good downside or loss, more power to you.

It's the insistence that there must be a loss, however small, however obscure, that loses me.

[Carlos Ferreira]

Scott,

Of course avoided costs are accounted for. It is a cost and benefit analysis, so you have to account for the benefit of healthcare versus the (predicted) cost of not having it.

[John Whitehead]

Odo,

Economics is a study of decision making. Understanding the costs of a decision, however small, helps us understand how decisions are made. For example, suppose the costs of a purchase are small and the benefits large (i.e., an energy efficient refrigerator) but consumers don't buy the item. Then we must scratch our heads and wonder why. It turns out that the the higher up front cost is given great weight by consumers. This suggests to a policy person that it might make sense to nudge consumers to purchase the energy efficient refrigerator (with a subsidy or something). Not thinking about the costs, "the refrigerator pays for itself," only leads us to think that consumers are dumb and we should force the decision on them. Consumers are dumb, but they do attach a heavy weight on current consumption.

[John Whitehead]

Odo,

"pathological," "sophism" etc.

Nice try, but, sticks and stones. Sticks and stones.

[John Whitehead]

Scott,

Google: "value of statistical life"

[odograph]

To be honest John, you've really lost me on this argument. I don't know what the point is anymore.

It seems important to env-econ.net that green jobs be "bogus" and that there be no "win-win" but the actual losses for things like efficient refrigerators or LED traffic lights are smoke in our fingers. Insubstantial.

That's why I've used the heavy handed terms, the pathological, the sophism.

I mean, did you ever find a good solid fault with efficient refrigerators or traffic lights? Something that you could hang your hat on?

If the only lifeline is a worry that maybe, somewhere, for some buyers, there might be an opportunity cost higher than the benefit is a pretty weak claim. It is hardly universal, as I've said, and more than that it is an appeal to missing data.

You don't show the particular opportunity or its cost, you just vaguely pose the possibility, to make what you really want to be a general case.

[odograph]

BTW, if it wasn't clear to the reader, remember that opportunity costs are possibilities rather than some fixed factor. If Jill and Jane both consider the same refrigerator (or traffic signal) they might have different finances, obligations, etc.,

They'd need to talk to their financial manager (or put on that hat), who might say "no, we don't' really have a better project than this right now."

At that point the hypothetical opportunity cost drops to zero. Indeed NOT DOING THE PROJECT would incur the opportunity cost.