Assorted links

by on January 6, 2013 at 2:47 pm in Uncategorized | Permalink

1. Charles C. Mann reviews Bernard Bailyn.

2. Becker and Murphy on whether we have lost the war on drugs.

3. How loud should religious announcements be?

4. A paper on lead exposure and crime, internationally (pdf), and some predictions here, including that Latin American crime will fall sharply.

5. Is it too easy to meet someone new?

6. What will induce nostalgia in 2033?

7. Is Social Security in worse shape than we think?

dearieme January 6, 2013 at 3:13 pm

“Historic declines in smoking and improvements in the prevention and treatment of cardiovascular disease are adding years of life that the government hasn’t accounted for.”
The smoking point is good, the heart/stroke point less so. Though there have no doubt been improvements in treatment, the great rise and subsequent fall in heart attack rates during the last 80 or so years is essentially unexplained (after you allow for the decline of smoking). If it’s not understood, predicting its continuation into the future is just a naive exercise in trend projection. It may work well; if those people who guess that the whole thing has been a matter of infection are right, then presumably there’s a lot of resistance to that infection around, so a continuing fall might well occur. Nobody knows.

” … their job is not to make statistical predictions. Yet the agency badly needs such expertise.” I’m mildly surprised to learn that this is outside the expertise of American actuaries: still, in a sense it’s outside everyone’s expertise if causes of changes aren’t known. Then there’s the potential intrusion of new phenomena: what will be the effect of antibiotic-resistant bacteria? Who knows?

Claudia January 6, 2013 at 4:18 pm

I think the actuaries might be surprised too, to learn these questions are outside their expertise. Here’s the relevant report from the SSAs Actuaries on the demographic assumptions: I recall from my time as an RA and working on SSA reform that there has been a debate for a while about the demographic projections. And I will note that they had some tech-ed up stats folks come in and do alternate projections. If you glance through the report there is more than one set of projections. I didn’t see anything new in the article, but some messages require repeating…Social Security has for a long time needed the political will to tweak it, still waiting.

Terri January 6, 2013 at 6:17 pm

No amount of statistics can predict the future

The whole article makes me chuckle. They are just trying to get the spotlight. Statistical methods can’t predict what medicine will come to the market in 5 years and what impacts it will have on mortality in the the projection period nor can it predict broad social trends so any statistical estimation of variation is moot. All you can really do is extrapolate existing trends and make assumptions (which is what SSA is already doing). The past is not the future so any methodology will be biased to some degree. I am very pro advanced statistics when it actually improves information but this is just the pretense of knowledge. Adding complexity to modeling just too make it more nerdy and more obtuse doesn’t really have a point.

Disclosure: actuary

Brian Donohue January 6, 2013 at 10:36 pm


Tyler, you need to pick your battles better. Social Security is a distraction from the orders-of-magnitude bigger issue of health care spending. And to the extent Social Security is an issue, mortality is a distraction. That article was lame.

If your position is that the trust fund is a fiction, you endorse the view that general revenues over the past few decades, and the low marginal income tax rates during most of that period, were supported by a regressive payroll tax that falls predominantly on the private sector.

Willitts January 6, 2013 at 11:12 pm

The point is that Social Security obligations are inextricably linked to obligations for senior citizen health care. The dollars that pay for health care increase the costs of Social Security. It doesn’t matter that one is “orders of magnitude” larger than the other when they both point in the same direction.

The SS Trust Fund is smoke and mirrors. The non-negotiable bonds are an obligation on future general tax revenues. Social Security benefits are an obligation on future general tax revenues (that just happen to be called something else). We’re paying for two obligations with one revenue stream. We’ve created an IOU for an IOU.

This should remove all doubt that Social Security was ever intended to be a self-funded system. Looking at the last generation is not the measure of the solvency of the system. If the government had done stress tests factoring in population booms and busts as well as business cycles, it would have been obvious from the start that insolvency was only a matter of when, not if. On the other hand, Social Security benefits were rather modest in the beginning and were NEVER INTENDED to be the sole source of retirement income. This Titanic wasn’t hit by a wandering iceberg, it steered into one.

Ricardo January 6, 2013 at 11:48 pm

“It doesn’t matter that one is “orders of magnitude” larger than the other when they both point in the same direction.”

If orders of magnitude in dollar terms don’t matter to you, I have a house I’d like to sell you.

Let’s be specific here: the latest Trustees Report projects SS expenditures will peak at 6.4% of GDP. These guys in the NYTimes article say that’s too low but their article says they project expenses “that are 0.66% greater of projected taxable payroll compared to SSA projections by 2031.” In GDP terms, it’s tough to see this anything more than a rounding error. On the other hand, Medicare+Medicaid spending will be 9.6% of GDP by 2037 according to the CBO’s long-term budget outlook.

Not only is that 50% higher than SS spending by the 2030s but it also represents a much more rapid growth rate in expenses between now and then.

Kevin January 7, 2013 at 4:03 am

You’re making a silly argument. When Social Security was introduced in 1935 who could have predicted that human lifespan would increase to the extent it did, fertility rates would go down, women would join the workforce, medical technology would become expensive yet effective, etc. If all of these variables had stayed constant funding Social Security and Medicare would be a piece of cake.

Andrew' January 7, 2013 at 8:06 am

It wasn’t possible to predict that defined contribution plans might be more actuarially robust than defined benefit plans?

Brian Donohue January 7, 2013 at 9:31 am


It depends what you’re trying to accomplish. Social Security accomplishes several things not covered by defined contribution plans, principally disability, death, and longevity insurance. Now it’s not 100% transparent, and there are, as I mentioned, other subsidies built into the system that you may take issue with, but the “social insurance” aspects of the system are very real and represent valuable “pooling” of these risks at the societal level. In my mind, one big pool is the most efficient for these risks. I think perhaps Hayek would agree, so long as “forced saving” doesn’t go overboard and require people to live more penuriously in the here and now than they need to. (c/f behavioral econ 101.)

For something like a third of current retirees, Social Security is basically the only source of retirement income, and for all but the very rich, Social Security is an important component of retirement security, so it doesn’t feel to me like the system is “forcing” too much saving for retirement.

This does leave the taxpayer as the backstop “insurer”, but I contend that the actuarial issues here are EXTREMELY tractable compared to open-ended things like healthcare- as I mentioned, a 20% benefit cut is likely on the cards as part of the solution.

Friedman is rightly concerned that every tax dollar has a bite taken out of it before a dollar of spending comes out the other side, but Social Security has a 77 year history of payng out 99% of tax dollars received as benefits, unlike virtually any other government scheme. I shudder to think what kind of bite insurers and Wall Street bookies would take out of a defined contribution approach.

And if instead, the defined contribution plan was some kind of “notional account” obligation of the US Treasury, you’d have faced all the same “tempting cookie jar” issues presented by the Trust Fund surpluses.

Cliff January 6, 2013 at 11:46 pm

And by “regressive” do you mean “highly progressive”? Higher income folks are taxed more and those taxes are used to subsidize lower-income people. Higher-income earners are truly taxed because they pay much more than they get in benefits, while lower-income earners get much more out than they put in.

Brian Donohue January 7, 2013 at 12:58 am


Assuming that the benefit promise is ultimately honored, yes, there is a “redistributive” element to Social Security, although even here, to the extent low vs. high income folks have different life expectancies (and they do), that is a mitigating factor (not to mention other subsidies, like benefits to non-working spouses.)

But honoring the promise means finding income (or other) taxes to repay the Trust Fund. If the promise is not honored in this way, then I believe “financing general operations via a regressive payroll tax” is an apt characterization of tax policy over the past three decades.

Brian Donohue January 7, 2013 at 1:03 am

And Cliff,

I don’t view ‘honoring the promise’ as current law benefits set in stone. Social Security benefits were already ‘cut’ by 15% for people born in 1960 or later with the 1983 law, by increasing the full retirement age from 65 to 67.

Increasing the full retirement to, say, age 70, would amount to an additional 20% cut in benefits- along with some other tweaking, this is about all the system needs to remain viable, assuming the Trust Fund is repaid.

Ray Lopez January 7, 2013 at 6:10 am

“Tyler, you need to pick your battles better” – rhetoric noted. TC does not pick battles, he simply throws everything and the kitchen sink at you. That’s one thing I like about his blog–it’s eclectic.

Andrew' January 7, 2013 at 7:56 am

Tell your bank that your credit card bill is orders of magnitude bigger than your mortgage payment. It doesn’t matter.

The problem is not just that SS is 1/4th or whatever the problem that unchecked medical spending is, both have already crowded out almost the entire government.

Brian Donohue January 7, 2013 at 9:45 am


I disagree. Tyler’s links are not, I trust, the kitchen sink. A chimpanzee could do that. We all have limited bandwidths, some (gasp!) discrimination is called for in these links.

I don’t want to sound dickish, though, cuz I think Tyl;er is a Force For Good in this universe.

Lion of the Blogosphere January 6, 2013 at 3:21 pm

Ultimately, Social Security is funded by the taxing power of the United States. The whole trust-fund thing is a fraud. We have to assume that a few decades from now, the people running the government will have some common sense about how much to pay out and to whom. I wouldn’t worry about it.

jdm January 6, 2013 at 3:51 pm

>”What will induce nostalgia in 2033?”

Fading memories of the Holocene.

Mark Thorson January 6, 2013 at 7:51 pm

Life before Google became the government.

Thor January 7, 2013 at 1:39 am

How easy it was to meet people!

Mark Thorson January 7, 2013 at 9:47 am

Google’s big data algorithms do a better job than “organic” matchmaking.

Go Kings, Go! January 7, 2013 at 1:48 pm

Being allowed to drive your own car.

Mark Thorson January 8, 2013 at 7:25 pm

But most of the fun driving in the Googletopia is when you find out where your car is taking you! It’s always someplace great! Someplace you would have wanted to go, if you had known as much about what you want as Google knows.

Go Kings, Go! January 7, 2013 at 1:51 pm

the Holocene

You expecting another Younger Dryas? The world cooled 15 degrees (Celsius!) in a single century, if my memory of Barry Cunliffe’s Briton Begins is correct.

8 January 6, 2013 at 4:07 pm

Marginally speaking, SS is a problem. More cash is going out than going in, which makes SS a burden on the government. If they don’t reform now, SS payments will be on the chopping block when the debt crisis hits.

Ricardo January 6, 2013 at 10:54 pm

From a political economy point of view, SS benefits a group of people who have high voter turnout and who make up a significant and growing percentage of the population. Social Security spending is around the same magnitude as defense spending and is smaller than Medicare+Medicaid spending — it will grow to be larger than defense spending in the coming years but will always be less than Medicare+Medicaid. I doubt there will be a “debt crisis” (which some people have evidently been predicting since the 1970s) but even if there is, it’s likely that spending cuts will be across the board and SS may avoid serious cuts because it is such a politically sensitive issue supported by a powerful group.

Careless January 6, 2013 at 4:19 pm

@5: The number of women in that comment thread calling “Jacob” a “douchebag” for, essentially, getting on with his life after his long-time girlfriends left him is… interesting.

Miley Cyrax January 6, 2013 at 4:19 pm

An answer to 6. can be sought from 7.: “…a $2.7 trillion buffer built in anticipation of retiring baby boomers — will be exhausted by 2033, the government currently projects.” However, I presume a kick-the-can-down-the-road measure will be implemented by then.

5. is a riot, with generous butt-hurtery and ad hominems directed at both the author and “Jacob” in the comments section as the cherry on top. Surprise, surprise–men will be lazy/incompetent (“He’d been called lazy, aimless, and irresponsible with money…”) and averse to commitment if they can still retain female sexual access, which (for some men) is facilitated by online dating.

Ray Lopez January 6, 2013 at 4:31 pm

@4- the old “lead and crime” nexus thesis, akin to the “lead pipes doomed Rome” thesis; if only it was so simple.
@5 – ““It was fairly incredible,” Jacob remembers. “I’m an average-looking guy. All of a sudden I was going out with one or two very pretty, ambitious women a week. At first I just thought it was some kind of weird lucky streak.” –> I know this feeling. Just move to Thailand or the Philippines or Malaysia. U Hand-sum Man! All Westerners look like Hollywood’s Sexiest Man Alive with Bill Gates money.
@7 – another SS scare story that won’t sink. SS will owe $$$ Trillion by Year XYZ. Alarmist, sound the alarm? No. They will simply extend SS age limits and cut benefits. Problem solved. Move along, nothing to see here.

Jonathan M.F. Catalán January 6, 2013 at 5:14 pm

Becker and Murphy mention that the drug war stunts development of methods to help addicts. They don’t directly mention it, I don’t think, but the complete decriminalization of drugs, including the legalization of their production, would also stimulate the development of new, better — including less addicting and less dangerous — drugs. It would also push drug prices down, making it unnecessary for people to go for cheaper, more addicting, more dangerous, drugs (e.g. cocaine to crystal meth).

Brian January 6, 2013 at 8:12 pm

// It would also push drug prices down, making it unnecessary for people to go for cheaper, more addicting, more dangerous, drugs (e.g. cocaine to crystal meth).

That describes the transition from Benzedrine to cocaine and from cocaine to crack cocaine but not the shift from cocaine to crystal meth. Banning over the counter sale of amphetamine (usually branded Benzedrine) made the cocaine market. Cocaine was more addictive and had worse side effects but was cheaper to grow and smuggle.

Then crack as a cocaine deliver mechanism made it cheaper and more dangerous to get the same high.

But methamphetamine is actually not more addictive or dangerous than cocaine. It’s just cheaper. Probably it’s a bit less addictive. And it’s still much worse for you than the simple Benzedrine and Dexedrine they banned in the first place. You can still get those better stimulants if you’re an upper middle class white with health insurance: just ask your doctor for Adderall.

Dismalist January 6, 2013 at 8:32 pm

What Becker and Murphy could have pushed much harder is Becker’s own idea [I think] to put an excise tax on currently illegal drugs. Initially, the tax could equate the future legal price with the current street price, or perhaps raise it a little. No increase in use, and all those resources devoted to turf wars and their interdiction turns into tax gravy!

My own axe is that most of this stuff didn’t become illegal until the so-called progressives appeared, around the turn of the 19th to the 20th century. Let’s call them “regressives”!

There, I feel much better.

Andrew' January 6, 2013 at 5:44 pm

4. 18 year lag! That’s what you call a leading indicator!

Brian January 6, 2013 at 8:20 pm

Mexico City, where the lead phase out started before 1985 and was widespread with much cleaner air before 1990, has already seen violent crime drops.

The country as a whole has seen murder more than double since starting to cooperate with US drug prohibition in 2006. But Mexico City in particular has been getting safer.

Cleaned up gasoline took a few more years to reach other urban areas in Latin America. The new president of Mexico has put known narco-gang associates in security positions so the US-style prohibition agenda is probably going away. That means we’ll never really see whether lead is the major factor in a reduction of violence in Mexico because it’ll be confounded with a wind-down in drug violence.

Edward Burke January 6, 2013 at 6:05 pm

6. Cash and currency.

Edward Burke January 7, 2013 at 12:25 am

And coin.

Margin January 6, 2013 at 6:32 pm

“Becker and Murphy on whether we have lost the war on drugs.”


I never waged a war on drugs.

Mark Thorson January 6, 2013 at 7:52 pm

They lost a war on drugs. My side won.

Enrique January 7, 2013 at 12:29 am

From a normative perspective, the “war on drugs” is a senseless and stupid enterprise, but the non-normative question is this: does drug-prohibition reduce the consumption of drugs? On that question, we don’t have a real answer (yet). Instead of providing premature speculation or an ideological answer to this question, why not wait and see what happens in Colo. and Wash. before we try to answer that question?

gabe January 7, 2013 at 11:28 am

Odd assumptions Enrique has.

I thought the normative question is “Does the war on drugs increase the power of the federal government and help in the war to destroy individual rights, while helping to fund black ops?”

Peter Schaeffer January 6, 2013 at 7:16 pm

We have lost so many wars of late… Drugs, sexual assault, physical assault, robbery, burglary, DUI, murder, home invasion, etc.. The police keep arresting the guilty and still the crimes continue.

Time to give up.

Cliff January 6, 2013 at 11:52 pm

Good job being purposefully obtuse. Policing drug “crimes” is a net negative to society, unlike those other things.

Frank Youell January 7, 2013 at 2:11 am


“Policing drug “crimes” is a net negative to society”

How sure are you? How much would addiction go up if crack was legal? Opium? Meth? Barbiturates? China de facto legalized opium in the 19th century. By some estimates 25% of the population was eventually addicted. How much did that cost the Chinese? More than stopping the opium trade?

Opium addiction was a material cause of the communism revolution in China. What did communism cost? It’s worth noting that the Chinese communists and their opponents have always agreed that drug enforcement is vastly cheaper than drug addiction. Perhaps more relevantly, they have been highly successful in suppressing the drug trade.

If you doubt this, let me suggest a career selling heroin in Singapore. It will be short.

MSL January 7, 2013 at 4:07 pm

@Frank Youell

People were dumber then. The “war-on-drugs” is foolish.

Marian Kechlibar January 7, 2013 at 11:56 am

The difference between drugs and all the other things that you mention is that drugs are “malum prohibitum”, all the other things are “malum per se”.

In other words, the vast majority of the population loathes the other crimes and will go to great lengths to suppress them. On the other hand, recreational drug use is seen as an unjustly suppressed, victimless, legitimate activity by huge and continually growing fraction of the population.

I would say that enforcing laws which criminalize an activity which is perceived as acceptable by a large part of the population is a) undemocratic, b) fundamentally immoral.

sourcreamus January 7, 2013 at 3:09 pm

If the portion of the population that would like to ban drugs is bigger than the portion that would like to legalize drugs than how is banning drugs undemocratic? It is by definition democratic. Also since when does fundamental morality depend on how many people like or dislike something? Was slavery moral because a majority of slave owners liked it? Your thinking is all muddled.

Why should all drugs be lumped together? Marijuana makes you eat doritos, bath salts make you eat other people’s faces, should not our laws acknowledge this difference?

Frank Youell January 7, 2013 at 5:23 pm

“The difference between drugs and all the other things that you mention is that drugs are “malum prohibitum”, all the other things are “malum per se”

Really? Throughout history, judgements on what is inherently wrong versus merely prohibited have varied quite a bit. Adultery and homosexuality have been punished by death. Theft has been state sanctioned numerous times. The definition of what constitutes rape has been all over the map.

“In other words, the vast majority of the population loathes the other crimes and will go to great lengths to suppress them”

If it’s not cost effective to suppress burglary or robbery why bother? Because a majority don’t approve? If a majority don’t approve of drugs, doesn’t that justify suppression?

Drugs are hardly victimless. Aside from ODs and DUIs, a welfare state and drugs don’t mix. The taxes needed to pay for drug driven welfare dependency are not exactly voluntary. If they state has the legitimate power to use force to collect taxes, why is it wrong to use force to prevent dependency?

Alex A. January 6, 2013 at 8:27 pm

Nostalgia 2033—
Themed parties for the aughts have to have original iPods and iPhones as accessories, and all the girls will wear Ugg boots. Angry Birds will be the official party game. Myspace and defunct social media company t-shirts will be a good ironic choice; costumes based on the rise of reality TV will be popular (Survivor, Big Brother, American Idol). Invitations will be written on the back of questionable mortgage documentation. Gift bags will include official Snuggies.

Alex A. January 6, 2013 at 8:35 pm

Addendum: Snack theme will be “Atkins/low-carb dieting”.

Bryan Willman January 6, 2013 at 9:15 pm

The real problem with the social security argument, and any argument like it, is that in the extreme both “trust funds” and ‘tax policy” are just accounting noise that ignore reality.

For any economy, there is some maximum total number of “non-productive” people it can support, one way or another. As an extreme illustration – if there are 100 retired people for every working person, social security won’t mean squat. And of course, “not working” and “non productive” are not related – observe any parent raising children, “retired” people who mentor and teach or just help maintain the local park.

But in the end, if the ratio of “people the employed economy must support” to “people who are active in the employed economy” gets too high, there will be a collapse, REGARDLESS of the accounting/taxation/trust-funds, or any other such nonsense.

So the question of the day is – what is that ratio, really? And what ratio will the US economy be able to tolerate over the next decades, really?

JWatts January 6, 2013 at 10:34 pm

“For any economy, there is some maximum total number of “non-productive” people it can support, one way or another.”

Yes, but you’d be surprised at the amount of people that don’t get that.

Brian Donohue January 7, 2013 at 1:21 am

Individuals contemplating retirement understand that working one or two more years, or working part-time for several more years, can have a dramatic impact on retirement finances.

And so for society (non-fallacy of composition.) Apocalypse…averted. It’s already happening, in this country anyway.

Ricardo January 7, 2013 at 3:42 am

In fiscal year 2011, Social Security spending was pretty close to defense spending. It will grow to be a bit higher in the coming decades but as a percentage of GDP, it will hover around 6%.

The ratio of workers to retirees is relevant to the question of how high payroll taxes need to be to finance SS but it’s a distraction from the more general question of whether SS spending is too high or is somehow unsustainable. The U.S. total fertility rate is near replacement so America is not Japan. The number of recipients of SS benefits is high but benefits per person are not particularly outrageous. That’s why the military — which employs far fewer people than there are SS recipients — shows up with a comparable level of spending.

And if you think that 6% of GDP for a government program is intolerable or will result in “collapse” but that 4% of GDP might be sustainable, I think you would be under an obligation to show why that is the case with some references.

Kevin January 7, 2013 at 4:32 am

This whole conversation about Social Security and Medicare is constructed on an infantile premise that seniors won’t organize politically to ensure that they have access to at least a minimal pension, health care, including heroic efforts, and long-term care. Anyone who doesn’t take this as a given is 1) young or 2) a fool. How are you going to keep them from voting? Or, really, to keep you and members of your generational cohort from voting?

Yancey Ward January 7, 2013 at 10:43 am

At one time, Japan wasn’t Japan either.

jorod January 6, 2013 at 9:54 pm

Is crime caused by lead exposure or exposure to the welfare state?

DocMerlin January 7, 2013 at 12:06 am

War on drugs? No, they didn’t lose the war on drugs. It was never about drugs, it was about increasing government and prosecutorial power. At that its been successful beyond anyone’s wildest dreams.

Yancey Ward January 7, 2013 at 10:44 am


gabe January 7, 2013 at 11:30 am

We have one that can see!

Thom January 7, 2013 at 3:03 pm

Yes. That $40B we spend on extra police every year is one of the biggest features to the politicians who seek to control us.

Jeff January 7, 2013 at 12:16 pm

I stopped reading the lead paper when I got to the chart that plotted gasoline lead per capita along with a crime rate 23 years later. Total gasoline lead would make sense, per capita gasoline lead does not. The concentration of lead in the air varies with the amount put into the air by burning leaded gasoline. It doesn’t vary because the number of toddlers changes. Unless, of course, you think every car exhaust was piped directly into the lungs of a toddler and held there until all of the lead was absorbed.

Jack January 7, 2013 at 8:14 pm

@4: The lead-crime paper looks to me incorrect: a classic spurious regression (google it). The authors cite Steven Levitt but miss entirely his contribution, which was to find and use instruments to establish causality, precisely when an analysis of time trends would be misleading due to the spurious regression problem.

Ricardo January 7, 2013 at 9:49 pm

It seems to me the whole point of the crime-lead link (as well as the apparent link between lead exposure and IQ) is very simple. It may or may not be a spurious regression but how many rigorously proven, relatively inexpensive interventions do we have to reduce crime or raise IQ? Not many. And unlike with other potential interventions, we do know something about the biology of lead exposure and how it affects the brain. If you want to establish causality, do an RCT of a lead removal program and that can clear up whether or not the regression is in fact spurious. The evidence is compelling enough to go to the next level.

Jack January 8, 2013 at 4:29 pm

Point taken. If there is a good medical-biological theory behind the claim, then I agree that such a naive regression is nonetheless useful and sufficient to justify funding a RCT to get a cleaner cause-effect result.

I am just in general bothered with data mining and Kanazawa-style random explanations.

Comments on this entry are closed.

Previous post:

Next post: