Assorted links


3. I noticed there were no subscription cards in a recent WQ I bought. In related news, GOOD has fired all its senior editorial staff. It's tough to make it as a thought journal these days.

So there's this thing called the internet, and on it are these sites called 'blogs'...

Wow, I wasn't expecting anybody to make that comment.

You're welcome?

Re 1: In addition to the list provided, there are all kinds of reasons dealing with the local physical environment that could lead to such improvements in life expectancy.

Perhaps there was a radical decrease in factories located in NYC over the past several decades, especially in or near the Bronx, which previously spewed airborne or waterborne carcinogens locally? Similarly fewer workers in nasty factories might lead to life expectancy increases.

Perhaps New Jersey is spewing fewer airborne carcinogens into NYC's air?

Similarly, I recall during the 80's that GE was harrangued about PCP's in the Hudson River...perhaps the NYC area had far more polluting factories than most of the country that no longer exist or which are far cleaner today?

Perhaps it has something to do with the water, e.g. maybe whoever is in charge of NYC's water provision discovered something horrible that they were doing, and instead of telling the public about it, they simply fixed it rather than risk a lawsuit somewhere down the line, which led to marked life expectancy increases?

Perhaps (some) Superfund cleanup sites that many of us mocked as expensive wastes of money actually had huge positive outcomes?

Perhaps the dramatic reduction in crime in NYC also dramatically reduced stress for locals, noting that stress significantly cuts life expectancy.

How homogenous are life expectancy data when contrasting different cities and regions across the US?

Crime may be the main factor. Specifically, reduce murder rates and traffic fatalities and life expectancies can shoot up. Folks dying under age 30 really bring the averages down.

Folks dying under 30 is a very quick way to bring the averages down although, even in large cities, crime isn't high enough to account for such a disparity.

Rather, I suggest that the single largest factor in determining life expectancy differences today is at work here, too: infant mortality. I would posit that perhaps mothers are leaving the city to have their babies in increasing numbers, causing a decline in those deaths being reported for NYC.

Also a good point

If anyone is leaving NYC to have a baby, they are probably well-off couples going to live in the suburbs. They have relatively low levels of infant mortality. Also, more and more people are choosing to raise children in the city over suburbs.

It's not so much a lower crime rate as the fact that people in NYC drive a lot less than people in almost all other parts of the country.

Life expectancy for a place with NYC with lots of immigration and emigration is incoherent. To make sense of it, demographers have to make a compromise. I think that the definition they use depends only on the infant mortality rate, and not on the birth rate.

Anyhow, while many people have always left the city to have children, the recent change is that more stay.

"Crime may be the main factor. Specifically, reduce murder rates and traffic fatalities and life expectancies can shoot up. Folks dying under age 30 really bring the averages down."

No, there's not enough murder in the US nor in NYC to make those kind of differences directly.

From the article:

"Although homicide rates dropped, they were never high enough to affect the life expectancy of millions of people on this scale. To increase the life expectancy of one million people one year, about 20,000 fewer 25 year olds would need to be killed. In 2008, roughly 60 people were killed per one million citizens. This isn’t the cause."

Ergo, the murder rate would have to be 500 times greater than it is for that to be the reason.

However, it is possible that NYC's crime rate drop led to an influx of relatively healthy people and the resulting gentrification pushed out a group of poorer less healthy people.

No, his math is definitely very, very wrong.

The overall death rate in NYC is known-- it's 6.3 per 1000 people, or roughly 50,000 people die each year. That is a very good number, better than the USA average.

The average age of death exceeds 25.

What he appears to be claiming is that if absolutely no one died in NYC for an entire year, which would only prevent 50,000 deaths, with an average age above 25 and thus fewer than 50 person years saved per death, then the life expectancy of NYC would rise by quite a bit less than one year, something like one-third of a year.

That can't possibly be correct.

This is just not true. You can try the math out yourself very easily . . a reduction of a couple hundred murders a year, in a city of 8 million, will have an immaterial impact on life expectancy.

About 50,000 people die each year in that city of 8 million. You're right that a "couple hundred" per year doesn't have that much effect. But the 1500 fewer from murder has some effect. The 6000 fewer per year from AIDS also has some effect.

I mean, the number of deaths per year declined by well over 10 percent among some causes that disproportionately affect the young. Surely that had to have a significant effect.

New York City in the period studied declined from around 70000 deaths per year to 53000 deaths per year. The entire life expectancy change results from those 17,000 deaths prevented. Of them, over 6000 per year were AIDS deaths, which skew young. Therefore they are the primary cause. There are secondary causes.

You can try the math out yourself. Unfortunately, the linked post does the math incorrectly.

Blasphemy around these parts, I know, but perhaps increased local government intervention had an impact.

#1: the graphs tend to jump around 1997/98. Is it just an older population dying out? Where's Steven Levitt when you need him...

Re: Number 1:

1. Average LE in Bronx, the "blackest" part of NY, is just shy of 78 years. The average LE of African Americans is around 73 years or so. Whatever is going on, I think we can assume that blacks in the Bronx are doing much better than blacks elsewhere.
2. This phenomenon isn't distributed evenly across the rest of the US. the New York City area did exceptionally well, including Conecticutt and New Jersey suburbs. But there's a huge part of . . "real America" that went backwards. And no, there isn't a demographic case to be made because its mostly happening white America ("real").
3. People in this thread are already going to great lengths coming up with contrived explanations that dance around evil gubnant's impact. Can't wait to see the next 50 responses.

The biggest jump was in the '90s. What government program occurred then?

What strikes me about that map is how often the declines in life expectanciy stop right at state borders, most specifically Oklahoma. Almost all of southern and eastern Oklahoma had declines, but they didn't slop over into Texas and Arkansas. Same kind of thing in Arizona and Nevada.

There's something not right about that map.

Pretty sure it's just a feature of the way the data was collected - namely, by county. I don't know for sure, but I'd bet that a quick Google search could confirm my guess that all those counties stop at state lines - thus, the declines stop at state lines, too.

@Ted Craig

Increased environmental regulation - especially in the late '80s (debatable if you want to count that as a program), making life more difficult for cigarette smokers (higher taxes, much more aggressive advertising, smoking bans), increased government spending on providing healthy meals (in NYC, for example, the local government provides free lunches for kids 18 and under), etc. It's possible that the jumps in life expectancy are not due primarily or even in any significant way from government action - I just don't think the evidence to support that view is that strong.

It's also possible that better security meant people felt safer spending time with their families in the park - more outdoor activity leads to better health results. That's just speculation, though.

I think your theories are pretty absurd. Subsidized school lunches are not going to have any impact on mortality for quite a long time if ever, unless people are dying of malnutrition which I doubt. Increased time in the park similarly does not pass the laugh test. Has the incidence of cigarette smoking in NYC been significantly different than elsewhere?

According to this, smoking is the leading cause of preventable death in NYC, or was a few years ago, anyway. Advances against smoking clearly and directly translate into better health benefits - much as the soda ban probably will.

I don't think that people can argue against the soda ban from a health perspective - it probably WILL help people be healthier. A more theoretical anti-paternalism argument could work, but really, I think the practical evidence bears out that people will live longer with a ban on selling 2-liter bottles of soda. You can question whether or not a few extra years is worth the nominal restriction on liberty (presumably, those who care would simply not drink (as much) soda regardless), but that's another argument.

As for the school lunches, it's not an issue of children starving to death in the streets. School lunches provide at least a baseline of nutrition, and proper nutrition means the body is healthier and better able to fight off infections, have stronger bones, etc. Plus, it's money that lower-income families don't have to spend on feeding their children.

Time spent outside is just a throw-in, and probably not a substantial cause. Nevertheless, spending time outside is crucial for vitamin production and general health otherwise - it's really not that absurd to imagine that, especially in worse neighborhoods, parents might not want to let their children play in the park if they were worried about their children's safety.

If school lunches make people generally healthier, 40 years from now that will show up in the statistics. Maybe you can see it immediately if you look closely, but I very much doubt this kind of life expectancy calculation extrapolates that much.

Smoking has more of an immediate effect, but still not very fast.

When you see something like this, its a data collection issue. In plain English, its a sign the data is garbage.

Even if this is truly a sign at the competence of New York City and state governments at keeping their citizens alive longer, the fact that the big jump occurred in the 1990s is another issue. There is really not much to see here.

A good place to begin would be to understand how life expectancy is calculated. (i) Is it life expectancy at birth? (ii) Do they mean the LE of current residents of NYC rather than of people born there? (iii) Having cleared up those facts, I repeat, how is it calculated?

For instance, if you do some social cleansing and push lots of the poor, old and infirm out of NYC and replace them with wealthy, healthy youngsters, would your LE figure improve? If so, and if there is a lot of population movement, it seems pretty silly to concentrate on frivolous government policies. Kelley's point about social cleansing of pregnant women is a similar argument.

I think you're right. For me it's more self selection than anything else.

Sick people leave NY. There can be a lot of reasons for that: cost and quality of health care, quality of life for sick people, NY real estate.

I think that one of the advantage to live in a big city like NY is that you can fin a high paying job and enjoy a richer cultural and night life. In exchange prices are higher especially real estate. I can imagine that if you're sick, you're not reaping the benefits anymore but you're still paying the cost. So it's sensible to leave.

Moreover, as you say, we can imagine that NY attracts more and more rich educated people who have typically a longer LE.

Perhaps this is also supported by the relative change of LE between the boroughs. As it becomes more expensive to live in Manhattan, less healthy people move from there to Queens because it is cheaper. Could look at the correlation between change in relative life expectancy between boroughs vs. cost of living.

Definitely a break in the mid 1990s. Some exogenous change in govt. policy ? This one comes to mind, but could be something else - we could say more if we looked at breaks down across demographic and socioeconomic categories :

HIV/AIDS treatments in conjunction with homicides. My hazy sense is that it was around the late 1990s that the nation as a whole and NYC in particular began to see improvements there.

"the USA—already trailing the world’s longest lived countries—dropped back further"

This is a bright red flashing BS light. The largest single influence on life expectancy is infant mortality, which is a self reported statistic with some extreme variations in the definition from one country to another.

Re the jump in the late 90's, I'd be the last to endorse large scale government intervention in anything, but note that S-CHIP began in 1997. It would be interesting to see if NYC S-CHIP programs differed significantly from those in other regions.

I'm curious to see the leading causes of death and if they changed over time. Wouldn't that be more informative than just a chart of life expectancy and speculation about the cause.

#3- I am surprised to see an econ blog reviewing a book about Edmund Spenser. Regarding the book itself, I don't plan to read it even though I have some interest in Spencer. A 600-page tome on this guy is just too long.

2. The proof begs the question. It's amusing, but you can't prove game theory is useful by using game theory.

#1 I am suspicious of the data. It seems like to large a gain though it could be due to a demographic change or maybe a change in drug culture.

Comments for this post are closed