Friday, September 27, 2013

Data link roundup (week of September 27, 2013)

The week's top data analysis links...
in honor of World Contraception Day.


The Guttmacher Institute is always my go-to source for information about contraceptive use. In honor of World Contraception Day, here are some striking facts about contraceptive use in the United States:
  • Far more married women (77%) than never-married (42%) use contraception
  • 99% of US women (age 15-44 and who have ever had sex) have used at least 1 contraceptive method
  • Contraceptive use is common among women of all religious denominations, including:
    • Nearly 90 percent of Catholics and Protestant women (who are currently having sex but do not want to get pregnant) currently use a contraceptive method.
    • Among sexually experienced religious women, 99% of Catholics and Protestants have ever used some form of contraception.


This map highlights states where unintended pregnancy rates are highest (Mississippi, Louisiana, California), and where annual public spending on unintended pregnancy exceeds $1 billion per year (in the states of California, Texas thanks, no doubt, to their large population size).

Source: Maps on the Web


Rising rates of premarital conception (and, likely, rising rates of premarital sex) accounted for the increasing number of premarital first births to women born between 1920 and 1949.

BUT for women born later (between 1945-64) increases in premarital first births were primarily attributable to declines in "shotgun marriages," according to new research published in the journal Demography.


In response to my post last week about the high frequency of August and September birthdays@DrDemography shared a great link... on how pregnancy due date may affect an infant's health. (For causality, think flu season...)


Following on last week's theme about the frequency of birthdays, this infographic shows not only frequency by birth month, but also by day... 
Source: IPH79 on Tumblr with data from the NY Times


Friday, September 20, 2013

Data link roundup (week of September 20, 2013)

The week's top data analysis links...


Late last week Slate published an excellent piece detailing the limits of longevity.

And speaking of longevity...
The Population Reference Bureau released the 2013 World Population Data Sheet and online interactive graphics.

Image source
And speaking of population growth...
The New York Times published an very non-Malthusian op-ed by Erle Ellis suggesting that there is no such thing as a "carrying capacity" for the earth.
(Thanks to @DrDemography for sharing the link.)

For what it's worth...
While I lean more toward Hans Rosling's "possibilist" attitude about population growth than toward Malthus' pessimist worldview, I also suspect Ellis has never seen the Star Trek episode The Trouble With Tribbles.

Also, on the blog Weeks Population, Dr. John Weeks published a succinct rejoinder to Ellis' op-ed.


Thursday, September 19, 2013

New data: 2012 American Community Survey

New data released today from the 2012 American Community Survey (and also from Tuesday's release of the Current Population Survey) show no change in median household income, poverty rate, and health insurance coverage in the United States (nor change across the majority of states in the U.S.).

While this may seem like a bad news story, 2012 is the first year since the the beginning of The Great Recession in which income did not fall and poverty did not rise. While the trends do not yet show growth, 2012 may be the inflection point the economy has been waiting for.

One key point of note, older Americans seem to be faring better - or at least recovering from the recession quicker - than younger Americans with substantially lower overall rates of poverty (despite a bit of an increase from last year for those age 65+) and improving homeownership rates.

For a concise summary of the new data:
William Frey, of Brookings, discusses findings from the 2012 ACS on Morning Edition. (Note: I love that reporter/interviewer Steve Inskeep refers to today's data release as "Christmas for demographers.")

Wednesday, September 18, 2013

(Not so) Popular opinion on population forecasting?

This week the myopic Globe and Mail reporters seem to think that if we cannot predict the population 50 years from now with complete certainty, we might as well hang up our hats and just stop trying.

That's their solution? Just quit. Ignore all data. Ignore all trends. Live like we don't have vital events registries and calculators, much less computers? I wonder if we should stop trying to predict the path of hurricanes, too, because, you know, we can't tell exactly where the eye will make landfall?

I'm flabbergasted by the implication that someone would prefer no information at all to a well-informed range of possibile outcomes. At first, I thought the article by Philip Cross must be an Onion link that I was mis-reading.

(I'll spare you the full force of my rant. Just envision a demographer unleashing a stream of profanities about willfully ignorant writers...)

What infuriates me most, I think, is the final paragraph wherein Cross writes:
"In the future, if the UN or Statcan insists on producing forecasts of population or the labour force, they should be obliged to include an easily-understood summary comparing their past forecasts and the actual outcomes, so people can assess for themselves the credibility of these exercises. It would also force statisticians and modellers to confront the flaws in their forecasts, something they seem blissfully unaware of and unaccountable for."
Perhaps Cross is blissfully unaware of the volumes and volumes of data, analysis, conference presentations, and peer-reviewed journal articles produced by demographers (including Stanley Smith , Jeff Tayman, David Swanson, Stefan Rayer, to name just a few) documenting exactly that: forecast error and forecast accuracy.

We demographers are not the ones who are blissfully unaware. In fact, there is an entire branch of demography devoted to improving forecasting techniques and measuring accuracy and error!

Cross, however, is blissfully unaware. Perhaps he should stop throwing stones until he moves out of his glass house.

Saturday, September 14, 2013

Why is September a common birthday month?

Note: This post is adapted from last year's post and includes new data and analysis...

Do you feel like you're spending a small fortune on birthday cards every September? Is your Facebook feed alight with birthday reminders this month? Have you noticed more birthday party invitations arriving in the mail in recent weeks?

If you have noticed a spike, you are not alone. There is a clear pattern of "birth seasonality" resulting from a "seasonal cycle in fecundability" documented in the scientific literature. Holidays and long winter nights are partly to blame for more birthdays in September, but human biology is at work as well.

In short, in the northern hemisphere, women are more likely to get pregnant in late fall and early winter than at other times of the year. As a result more births occur late summer and early autumn, a trend displayed clearly in the chart below with birth data from the United States in 2010 and 2011. (Note: In the southern hemisphere, the seasonal peak occurs about 6 months earlier.)
Source: U.S. Centers for Disease Control and Prevention and author's calculations
In the chart above, you may notice that March 2010 births appear high, but consider that February is a short month, and March is a 31-day month. August also has more days than neighboring month, September.

To correct for this, we can estimate the average number of births per day of the month. With this adjustment the seasonal pattern becomes even more pronounced.
Source: U.S. Centers for Disease Control and Prevention and author's calculations
Correcting for the number of days per month shows September as the clear leader for births in 2010, as it is in most years. But 2011 shows a slightly different pattern - August leads September by about 75 births per day.

The lower number of births in September 2011 may, or may not, be a consequence of a massive blizzard that shut down transportation along much of the east coast in December 2010...

Friday, September 13, 2013

Data link roundup: Higher Ed (week of September 13, 2013)

The week's top data analysis links...
This week's theme: Higher ed, higher earnings?


That higher education leads to higher pay is a belief so fundamental to public policy and private decision-making that it's rarely questioned. And, indeed, on average higher levels of education correlate with higher employment and earnings levels (at least according to the latest data from the U.S. Census Bureau).

But is all education created equal? Does a higher degree always lead to a bigger paycheck? Indeed, the story is more complex...

First, The Economist explains why it doesn't (always) pay to get a Ph.D.

Second, this week analysis published by Anthony Carnevale, an economist at Georgetown University, shows that not all majors are created the same (at least not in earnings potential).

Graphic source: National Public Radio


Child Trends published a summary of the demographic and socio-economic characteristics of the high school graduating class of 2013. Highlights include:

  • 89 percent are covered by health insurance
  • Nearly three quarters have experienced physical assault
  • 68 percent will go on to college
  • 35 percent volunteered at some point during the year
  • Less than half get the recommended amount of physical activity
  • 20 percent watch 4 or more hours of television per weekday


Source: Picking a college major XKCD