Let’s try for a little perspective. The BP oil spill threatens to cause something like $10 billion worth of damage. That’s pretty bad. By contrast, an extra trillion dollars worth of federal spending threatens to cause something like $300 billion worth of deadweight loss (that is, underproduction due to tax avoidance and disincentives to work). That’s 30 times worse. How is it that so much angst about the former seems to be coming from people with a history of shrugging their shoulders at the latter?
Both $10 billion and $300 billion are extremely rough guesses, but the $300 billion figure comes from the widely cited estimates of Harvard’s Martin Feldstein, according to which a one dollar tax increase triggers about 30 cents in deadweight losses. Since a trillion in new spending means a trillion in new taxes (either now or in the future), we get $300 billion in deadweight loss.
Of course $10 billion worth of oil-related damage is still big enough to be worth a goodly dollop of angst. But keep these things in mind:
- Occasional disasters are part of the cost of using fossil fuels. This might be a cost worth paying.
- On the other hand, even when a cost is worth paying, it’s still better to pay less. And this particular cost is likely to be excessive, both because limited liability inspires recklessness and because, in a sane world, we’d be drilling in the Arctic rather than a mile underwater near a population center. (Charles Krauthammer has made the latter point more elaborately.)
- On still another hand, excessive as the cost may be, it’s still piddling compared to the deadweight loss of taxation. Destroying $10 billion worth of what we’ve already got is not nearly as bad as destroying in advance $300 billion worth of what we could have had.
Now maybe $300 billion in deadweight loss is a price worth paying for the benefits of ObamaCare and the stimulus package—just as maybe $10 billion in damage is a price worth paying for the way we use energy. And maybe with wiser policies we could have slashed one or both of those costs. But one of them is thirty times the other, and if you’re looking to be justifiably angstful, that’s probably your better bet.
From my understanding, the $10B is direct damage and does not factor the long term losses like employment and fish-kill. But I may be off on that.
The bigger issue is where you come up with the $300B. You say it comes from, “Harvard’s Martin Feldstein, according to which a one dollar tax increase triggers about 30 cents in deadweight losses”.
But having read the article you cite (and come on, 1976? the research was done in 1976!) it very explicitly does not say what you are reporting it to have said.
Feldstein’s findings were that for every dollar gained by an incremental income tax COMPARED TO A LUMP SUM TAX there was a deadweight loss. The loss was accounted for by changes in behavior such as buying a home instead of renting and demanding pay in the form of employer healthcare. He’s comparing the differences of income tax vs. other forms of taxes.
This has come up before so I have to ask it again … would you accept this level of academic rigour in a paper turned in by one of your students? The research is dated and obscure and has been misrepresented, fundamentally, in a way that supports your position.
Tell me how this is just a case of inept research and not a matter of willful academic dishonesty. Or start spinning.
i have a problem with your second bullet. i live near the arctic.
it is true that a giant portion of alaska’s revenue is directly or indirectly a result of oil drilling. it is also true that spilling oil onto pristine wilderness, even if it is sparsely populated is an extreme cost.
the herring fishery in prince william sound is gone from the valdez spill. true. its only been 20 years…
you can still walk down to the beach and pick up a rock only to find a pool of oil underneath it.
the lower population density here seems to lessen the impact of the valdez spill only because less people have to see it. (bastiat’s unseen cost?)
another factor is that the people most affected by drilling here are the natives. i believe the arctic slope regional corporation is the wealthiest native corporation, but almost exclusively the oil industry is not employing natives or even non-native alaskans. while i am making no attempt to speak for natives, i did get a chance to hear a couple call up our senator the other day while he was taking calls on a radio show and they seemed to express dismay about some land-use decisions that they didnt feel included in.
even without accidents, i think one could make an arguement that the costs of drilling in the arctic outweigh its benefits.
have i seen mr. krauthammer on the o’ liarly factor?
why dont we want to import more oil and/or consume less?
i completely agree with what i have interpreted to be the point of your post. your third bullet is spot on.
i think that the 30% ‘inefficiency’ figure of the tax dollar might even be a little low. ive never seen a book in the private sector with pages that say ‘this page intentionally left blank’.
drat. my comment vanished. =[
i disagree with your second bullet. having less people see the oil only makes it easier to be one of bastiat’s hidden costs.
i completely agree with the third one.
The oil damage is visible and approximately measurable. The extent of the deadweight loss is speculative, contraversial and uncertain. It is also unavoidable, because the regressive lump sum tax is not politically possible.
Limited liability is likely to make this sort of event more frequent. The $75 million maximum cleanup cost is paid for by $0.08 / barrel tax, I believe, shifting the burden from producer to consumer. USA consumption is about 20 million barrels a day, or roughly $11,680 million in this tax since its passing in 1990. Enough to pay for over 1000 such incidents.
Apologies – temporary inability to count zeros – that is enough to only pay for only one such incident.
Benkyou: Yes, I did assume that most federal spending will continue to be funded by income taxes. Also that if the deadweight loss from an income tax exceeds the deadweight loss from a lump sum tax by 30 cents, then the deadweight loss from the income tax must be at least 30 cents. Which of those two assumptions did you find inept and/or dishonest?
Harold: It is also unavoidable, because the regressive lump sum tax is not politically possible.
The deadweight loss is avoidable by abandoning the spending program, just as the oil slick is avoidable by abandoning the use of fossil fuel. Since we’ve gotten along without this new spending till very recently, the former seems far MORE avoidable than the latter.
It is amazing that almost all of the rich countries have such high taxes.
The deadweight loss is so great that all the rich countries should be poor.
Inside the US, rich states have higher tax rates than poor states. Again, the deadweight loss should make states like Alabama to be so much richer than New York City.
Does it even enter your mind that things are more complicated than you wish they would be?
“The deadweight loss is avoidable by abandoning the spending program, just as the oil slick is avoidable by abandoning the use of fossil fuel. Since we’ve gotten along without this new spending till very recently, the former seems far MORE avoidable than the latter.”
You could just as easily say that we could abandon new drilling sites because we’ve gotten along without the new oil till very recently.
Why are you comparing only some spending programs to all fossil fuel consumption? As an economist, you should know that the marginal spending programs and the marginal oil drills are of most importance.
”
It is amazing that almost all of the rich countries have such high taxes.
The deadweight loss is so great that all the rich countries should be poor.
Inside the US, rich states have higher tax rates than poor states. Again, the deadweight loss should make states like Alabama to be so much richer than New York City.
Does it even enter your mind that things are more complicated than you wish they would be?”
This assumes that rich states and countries are otherwise equal to the poorer states and countries, which is very wrong.
Henry:
You could just as easily say that we could abandon new drilling sites because we’ve gotten along without the new oil till very recently.
Why are you comparing only some spending programs to all fossil fuel consumption? As an economist, you should know that the marginal spending programs and the marginal oil drills are of most importance.
Point well taken.
“The deadweight loss is avoidable by abandoning the spending program, just as the oil slick is avoidable by abandoning the use of fossil fuel. Since we’ve gotten along without this new spending till very recently, the former seems far MORE avoidable than the latter.”
Of course if the spending were not necessary, then abandoning it would be desirable. If the spending is necessary, then the deadweight cost is unavoidable. Have you just fallen into the same trap you accused Obama of yesterday? The spending program is beneficial or not, whether paid for by lump sum or income tax.
Harold: Steve covers this.
“Now maybe $300 billion in deadweight loss is a price worth paying for the benefits of ObamaCare and the stimulus package—just as maybe $10 billion in damage is a price worth paying for the way we use energy.”
I think I get the point – If we knock some percent off a large bill, we save more than if we knock the same percent off a smaller bill, so lets focus on the larger. However, I think there was just about as much debate and coverage of the Government spending before and after it happened as any normal person could stand. One hopes the legislators or their advisors are aware of these well documented costs. I feel sure there were enough people opposed to the measures willing to point it out to them. If Obama wishes to continue to a second term, then he will presumably have to deal with them.
The trouble is, the size of the bail-out was so huge it makes everything seem trivial by copmparison.
The deadweight loss is distributed over 300 million citizens, and will be spread over several years, depending on the maturity of the bond used to issue the debt, so the impact might be $100 per person per year. As Benkyou points out, the indirect impact of the oil spill will probably be much higher than $10B, and will be relatively much more concentrated. Fishermen, seafood distributors, and coastal recreation providers are and will be much more affected than, say, me.
@Steve, your statement “as maybe $10 billion in damage is a price worth paying for the way we use energy” ignores the possibility that spills like this are preventable. That’s like suggesting back in the ’70s that 2,000 to 3,000 airplane crash fatalities a year is just the cost of efficient travel, and that we shouldn’t bother to improve air safety. As this WSJ article (http://online.wsj.com/article/SB10001424052748704026204575266560930780190.html) points out, BP and their contractors circumvented federal regulations, and it’s possible that effective enforcement might have prevented this disaster. However, enforcement requires taxes. Sometimes taxes and regulation act as a form of insurance.
Al V.: Steve, your statement “as maybe $10 billion in damage is a price worth paying for the way we use energy” ignores the possibility that spills like this are preventable.
Right. Which is why I followed up with the statement that “even when a cost is worth paying, it’s still better to pay less. And this particular cost is likely to be excessive.”
You seem to be mistaking human emotion for economic analysis. I don’t think anyone is “angsting” over BP’s direct costs to clean up the spill (except maybe shareholders and so far they’re suing over other issues). The “angst” is over the destruction of livelihoods for tens of thousands of people, the potential loss of entire ecosystems, the deprivation of opportunity and other human factors. Attempting to lump all that together under a careless $10B label is just grotesquely missing the point.
Alan Wexelblat: I’d seen $10 billion as an estimate of the total damage. Have you seen a larger estimate somewhere?
Even if the figures are correct, and they seem about the right magnitude, isn’t this like comparing the 3000 fatalities of 9/11 to the 40,000 traffic fatalities that occur each year? The comparison does not mitigate the horrificness (if there’s such a word) of the event.
Well, we all know why we have this angst differential:
What is seen (oily pelicans) versus what is not seen (invisible deadweight loss). That, plus economists are not currently held in high repute.
Didn’t you post a link to Bastiat’s “Seen and Unseen” essay recently? That’s the answer. A duck caked in oil is seen. The deadweight loss is unseen.
Steve, I understand your point but this question seems a bit silly.
First, the proponents of Obamacare backed it because they explicitly believed that the benefits of the package outweighed the cost. You may not have noticed but there was also a significant contingency of folks who did not feel the benefits outweighed the costs. Those folks seemed to have a bit angst.
Second, the opponents of offshore drilling have been arguing for years that the benefits do not outweigh the costs. Indeed, their argument was based on the possibility of an accident even by a “well run” operations. There are now allegations that BPs operation was not well run.
If it was not worth it to explore coastal drilling even under the assumptions of a well run operation then certainly folks will be upset that drilling was done by a poorly run operation.
In short the sets of people upset about these two events are in large part DEFINED by their estimates of the benefits. It should not be unusual then to find consistent difference in implied benefit calculations.
On deadweight social loss: Loss compared to what?
As far as I can tell, the Feldstein article measures economic outcomes of a world in which taxes are raised via an income tax vs. a world in which taxes are raised via an otherwise unspecified “lump sum” tax. It is unclear to me how that lump sum tax would work, or what value Feldstein assessed to the consequences of implementing such a tax. Surely Feldstein wasn’t suggesting that a lump sum tax has no economic consequence. After all, every tax has marginal consequences at some point; even a head tax will influence how many children you choose to have.
Ironically, the “perfect” tax – the tax that leads to the least distortion of economic behavior — is a tax based on perfect discrimination. Note that in this blog post Landsburg raises no arguments against Obamacare and the stimulus plan; rather, he merely raises arguments against financing that spending through an income tax. He has raised no objection to, for example, financing those programs by simply appropriating lump sums from people.
By the way, the summary of the Feldstein article concludes, “Repealing the 1993 increase in tax rates for high income taxpayers would reduce the deadweight loss of the tax system by $24 billion while actually increasing tax revenue.” Since a recent article in the National Review discounted the idea that tax cuts pay for themselves, I’d be curious to get Landsburg’s take on this.
nobody.really: Note that in this blog post Landsburg raises no arguments against Obamacare and the stimulus plan; rather, he merely raises arguments against financing that spending through an income tax. I haven’t even done that. I’ve just pointed out that Obamacare and the stimulus plan cause far more collateral damage than oil drilling does, so that I find it curious that so many are so much more distressed about collateral damage in the latter case.
By the way, the summary of the Feldstein article concludes, “Repealing the 1993 increase in tax rates for high income taxpayers would reduce the deadweight loss of the tax system by $24 billion while actually increasing tax revenue.” Since a recent article in the National Review discounted the idea that tax cuts pay for themselves, I’d be curious to get Landsburg’s take on this.
Surely the revenue effect of a tax cut depends on the tax. Feldstein has thought hard about the 1993 tax increases and I haven’t, so it’s probably his take you should be curious about, not mine.
Perhaps I’ve misunderstood.
The only “collateral damage” Landsburg identifies is the damage of a 30% “deadweight social loss” based on Feldstein’s estimate of the distortions arising from using an income tax rather than a lump-sum tax. Because I don’t see where Feldstein’s analysis attributes deadweight social loss to government spending per se, I don’t see where Landsburg supports such an attribution.
Steven— The part where you put forward 34 year old research on a behavioral study as authoritative. Or where you attribute the deadweight to “underproduction … and disincentives to work” when that was not the conclusion of the research at all.
Benkyou: First, 1995 was not 34 years ago. Second, what I said was “Both $10 billion and $300 billion are extremely rough guesses, but the $300 billion figure comes from the widely cited estimates of Harvard’s Martin Feldstein, according to which a one dollar tax increase triggers about 30 cents in deadweight losses.” Surely this is nobody’s idea of putting forth a paper as authoritative.
So which are you: inept or willfully dishonest?
Wow. Lots of debate on this one.
I have to side with Steve here. It’s curious that people get so worked up over the decimation of vast amounts of fragile wildlife in service of our oil addiction. These people really need to just chill out and take a relaxing drive!
Plus, everything can be measured in dollars or barrels of oil, right?
Steve,
I like the way you think. The damage done to the economy just by the dead weight loss from the taxes necessary to pay for all the additional spending completely outweighs the economic damage done by the BP oil spill. Yet, we don’t see people frothing at the mouth over the dead weight loss of taxation. As someone else had mentioned above it is the seen verses the unseen.
Seen versus unseen may play a role, but IMO it is also unnecessary versus necessary. Steve may think there is too much government spending, and for that matter so do I, but guess what–we aren’t the deciders. The voters are. And if the voters want the spending, taxes are the consequence, and taxes need to be levied on the basis of ability to pay, for practicality if nothing else. So the DWL of taxation is a consequence of what we want to do, like the 40,000 traffic deaths each year, whereas the destruction of the ecosystem of the Gulf of Mexico is straightforward carelessness, just as the 9/11 deaths were straightforward malevolence.
Neil: Yes, I agree that necessary versus unnecessary has a lot to do with it. But occasional spills of this magnitude might in fact be a necessary consequence of using oil, and in the other direction the deadweight losses associated with the income tax might be unnecessary if we opened our minds to more efficient tax policies. So while the necessary/unnecessary distinction is important, it cuts both ways.
Oh snap. You cite an abstract for a working paper published in 1995 and on that very page it tells you the works origination “Published as “The Income Tax and Charitable Contributions”, Econometrica, Vol. 44, no. 6 (1976): 1201-1222″.
Did you even read the entire publication? Either of them? Or the most recent version published in 1999? Why would you fail to mention the parts where he very plainly explains that his deadweight estimates apply only to those with very high labor incomes. That people making less will incur little or no additional deadweight from a rise in income tax rate. And that the deadweight is a result of rise in availability of tax differed and non-taxed labor income (like employer healthcare).
Or how about the fact that his calculation is an admitted prediction based on (and this is a precious term for someone who calls themselves a scientist) “empirically estimated” figures. And how about investing the 10 minutes of research that shows his conclusions to be empirically false.(Feldstein, Martin. 1999. “Tax Avoidance and the Deadweight Loss of the Income Tax.” The Review of Economics and Statistics 81(4): 674-680.)
Like the 1993 rise in taxes under Clinton did not result in a net loss of revenue. Tax revenue rose steadily through the 90s. Where as in response to drastic tax cuts under W. Bush during the 2000s tax revinue often fell substantially. (http://www.cbo.gov/budget/data/historical.shtml).
You took a paper with clear self-defined limitations on its applicability. It says that a rise in incremental tax can have UP TO a 30% deadweight cost. And you apply the maximum end of the spectrum to a case that is clearly outside of the intent of the original paper.
If one of your students answered a question with “this answer is an extremely rough guess based on a 10 year old paper, based on 30 year old research, that doesn’t really say everything that I claim it does, nor support my argument in anything other than a tangential sense” how much red ink would you commit to that page? Or would it depend on whether that students answer (regardless of the quality) supported your pre-determined ideology?
And when you list something as a lone source for your argument you are putting it forward as authoritative. This is remedial rhetoric 101. If it was not authoritative then it does not belong as a source. If it is not authoritative then your argument is not authoritative and if that is the case you are wasting peoples time. If all you have is a source that is not authoritative then you have a weak argument and we deserve better research. If you cannot find better research to support your position than it simply may not exist or your position may be wrong. The BLS numbers, in this case, seem to indicate the latter.
My scholarship is thorough, I have put forward no information that I know to be false. I have supported my position (that this paper does not support your conclusions) with strong arguments.
Benkyou: I believe the problem here is that you are what is known as a “fucking idiot”. First, you have taken at face value a footnote that says a paper written in 1995, and containing references to approximately 30 papers written in the 80’s and 90’s, was in fact published in 1976. A more temperate person might have entertained the possibility that the footnote (appearing not in the paper but on a page pointing to the paper) is a mistake. Second, I didn’t reference the part where Feldstein says these estimates apply only to those with very high labor incomes partly because there is no such part, and partly because the eight pages of detailed calculations in Section 4 proceed from the clearly stated assumption that we’re talking about across-the-board tax hikes. Third, your “thorough scholarship” failed to turn up the fact that the 30% estimate, far from being the “maximum end of the spectrum”, is actually the low end estimate, with more realistic estimates of 72% when tax increases are across-the-board and 200% when they are concentrated at higher income levels. 72 and 200 are both substantially larger than 30. Fourth, you’ve apparently still failed to twig to the fact that if the deadweight loss from an income tax exceeds the deadweight loss from some other tax by x%, it is legitimate to conclude that the deadweight loss from the income tax is at least x%. And fifth, your habit of being at your most abusive when you are most wrong has become even more tedious than your habit of repeating the same errors over and over and over and over and over and over again no matter how many times they’ve been explained to you. I’ve pretty much had it.
Steve said: “Benkyou: First, 1995 was not 34 years ago. ”
The link says:
“*Published: Published as “The Income Tax and Charitable Contributions”, Econometrica, Vol. 44, no. 6 (1976): 1201-1222. Published as “The Income Tax and Charitable Contributions: Part I – Aggregate and Distributional Effects”, NTJ, Vol. 28, no. 1 (1975): 81-100.”
You have to pay to get the full text, but it looks rather like it is based on a 1976 paper.
Steve says “widely quoted”, which seems to intended shorthand for “widely accepted”. We saw earlier that a respected economist’s estimate of the “cost” of the stimulus package was based on almost laughably biased estimate of the “multiplier”, so it is possible for these things to be got wrong, even if they are widely quoted. I have no idea how widely accepted this finding is, possibly economists debate over it endlessly. I think it likely that it is accepted that there is some deadweight cost to taxation, so the question why do we not display more angst over the stimulus package and Obamacare, when we do over the lesser costing oil spill?
There is a very simple reason – the deadweight cost is an issue relating to taxation generally, and has nothing specifically to do with the issues mentioned. We would have to devote 30% of “angst” to each and every government measure. The fact is we have these debates about how we fund Government separately. Taxation occurs constantly and continuously – it would be absurd to raise these issues about one specific tax raising mesasure. The oil spill is a unique (so far) event. It makes complete sense to debate the issues relating to this event.
Steven has also conflated the stimulus package and Obamacare, with Obamacare put first. This creates the erronious impression that Obamacare is the bulk of the assumed $300 billion.
Harold,
1. Steve is referring to this paper: Feldstein, “Tax Avoidance and the Deadweight Loss of the Income Tax,” Cambridge, MA: NBER Working Paper W5055, March 1995.
2. The BP spill is just as unique as Obamacare and the Obama “stimulus” package.
3. There have been other studies such as, Charles Stuart, “Welfare Costs per Dollar of Additional Tax Revenue in the United States,” American Economic Review, June 1984, that show a dead weight loss of over 50% tax revenue and Charles J. Ballard, John Shoven, and J. Whalley, “General Equilibrium Computations of the Marginal Welfare Costs of Taxation in the United States,” American Economic Review, March 1985, that show a range between 15% and 50% of tax revenue.
4. All studies, however, have shown that taxes impose a deadweight burden on society. Steve, in choosing the Feldstein paper, has reference one of the most comprehensive studies and one that also falls somewhere in the midpoint of other studies that have been conducted.
5. Steve has, in my opinion, made an excellent observation that the deadweight loss of taxation from the additional spending under Obama completely outweighs the economic damage that is being done by the oil spill. If you would like to criticize Bush for the additional tax burden he had imposed during his years with runaway spending, such as the prescription benefit entitlement, I would welcome such criticism. But don’t stick your head in the sand when your guy is running the show.
While we’re on the subject of things worse than the Gulf Oil spill, so is Christmas. Much of the holiday spending is on gifts for others. At the simplest level, giving gifts involves the giver thinking of something that the recipient would like—he tries to guess her preferences, as economists say—and then buys the gift and delivers it. Yet this guessing of preferences is not easy; indeed, it’s often done badly.
In 1993 Joel Waldfogel, then an economist at Yale University, sought to estimate the disparity in dollar terms. His results put the dismal in the dismal science: on average, a gift was valued by the recipient well below the price paid by the giver.
The most conservative estimate put the average receiver’s valuation at 90% of the buying price. The missing 10% is a waste of resources that could be averted without making anyone worse off. In other words, if the giver gave the cash value of the purchase instead of the gift itself, the recipient could then buy what she really wants, and be better off for no extra cost.
If the results are generalized, a waste of one dollar in ten represents a huge aggregate loss to society. The study suggests that in 1993 America, where givers spend ~$40 billion on Christmas gifts, $4 billion is being lost annually in the process of gift-giving (actually the estimate was between $4 & $13B but I don’t want to be accused of inflating the figures). Add in birthdays, weddings and non-Christian occasions, and the figure is out of control. Once this oil spill has paid for, it’s done. This Christmas thing just goes on and on and on.
Clear water, functional ecosystems, and Christmas presents are for inefficient sentimental fools.
Here’s the study:
http://graphics8.nytimes.com/images/blogs/freakonomics/pdf/WaldfogelDeadweightLossXmas.pdf
Tom Dougherty,
The programs are not “unique” in the issue under discussion – that they lead to deadweight costs. The oil spill is I believe the only ever deep sea blowout leading to huge costs. That makes it unique. Obamacare is the only federal program leading to deadweight costs? No. That makes it not unique.
Thanks for the extra information regarding deadweight costs. As I said, I think we can take it there are *some* deadweight costs to taxation of whatever form. Whether it is 30% 2% or 100% is perhaps a red herring. We are comparing the costs of an accidental (negligent?) oil spill with two out of hundreds of federal programs. Why pick on theses two? The reason given is that they are occurring at approximately the same time. Why not pick on the deadweight cost of military spending, or roadbuilding, or the oil tax to pay for the liability fund, or policing, or prisons? Oil spills are the unavoidable consequence of oil drilling. Taxation is the unavoidable consequence of Government spending. The true comparison is not with Obamacare and stimulus package, but with taxation, or spending.
Suddenly the argument looks much stronger – why all the angst over $10 billion oil spill when the deadweight cost of taxation is about $800 billion each and every year? Or is it only $80 billion? That is the debate worth having, and the media does not have it. By picking on just 2 of the hundreds of federal programs, the argument is diverted into the benefits of those policies, which has already been discussed extensively.
Steve may be right to raise the issue, and try to make “seen” some of the unseen costs, but in my opinion wrong to single out just these two programs, which are not different in kind from hundreds of others.
Steven, to answer your initial question, people can get their heads around an emotionally-charged disaster like this oil spill, of which we have some experience. Most will have no idea what you are talking about with the “deadweight loss”, especially put in those non-layman terms. I, for example, had to look it up. haha I get it, but at the same time, I think this is still clearly an example of comparing apples and oranges.
Remember that article in the New York Times about regulation for the nanny market? And how you complained it focused solely on the benefits and not the costs?
You did the same thing here (in reverse), and-as an economist–you should be held to a higher standard! You’re an experts at reminding people that everything involves trade-offs. Minimum wages raise living standards, but cause unemployment. Taxes cause deadweight losses but give us … (lower mortality, peace of mind, more equity?).
(You did admit “maybe $300 billion . . . is a price worth paying for the benefits . . .” in the last paragraph. But that’s burying the lead.)