Know Your Source: Censuses Go Door to Door for Crucial Health Data
Every 10 years, the government wants to ask you some questions.
How you answer and how many of you answer makes a world of difference in how accurately researchers and policymakers are able to discern health trends and react appropriately.
I wrote about vital statistics in my last post and noted some of the limitations around information gleaned from birth and death records. One of the ways health researchers fill in the gaps left by those limitations is by using data gathered from censuses.
So what is a census?
It’s Christmas season right now, and it’s worth remembering a story many of you may have heard in church. Mary and Joseph were traveling to Bethlehem to be counted in a census ordered by Caesar Augustus, the Roman emperor at the time. That would have been more than 2,000 years ago. Since then, one big improvement in methodology is that in most countries the census is taken by government workers going door to door or through the mail. There still are examples of people having to travel long distances to be counted, but they are the exception.
Censuses count every person and each set of living quarters. They gather information about age, gender, education, occupation, religion, disability, race, and a variety of socio-economic factors. They pull together information about housing, access to water and power, and the availability of essential services.
There is such a wide variety of information that can flow from a census that I could spend several posts on the topic, but here’s just one example. While birth certificates are a snapshot of each person born, a census can provide a motion picture of family dynamics and overall fertility trends in a population. They can capture, among other things:
A. The number of children ever born in a household or to a person.
B. The number of those children still alive.
C. The date of birth of last child born alive.
D. The age, date, and duration of first, second, or third marriages.
E. The age of a mother at the birth of her first child born alive.
F. Household deaths in the past 12 months.
This type of information is what’s used to make estimates for birth rates and death rates in communities where good vital registration data are lacking.
To turn census information into useful data requires many steps. Those handwritten forms you filled out have to be converted into an electronic format. Then they go through a correction and validation process where the scanned forms are double-checked and inconsistencies are fixed. The actual analysis is the final stage.
But that doesn’t mean censuses are perfect sources.
For one, they are infrequent. Ten years is an incredibly long time. My brother Andy, who wrote those posts about football injuries last week, is 10 years my junior. There are times when I feel like he and I had roughly the same childhood because we both grew up in Great Falls, Montana, and had the same parents. But so much happened in those 10 years when he was not yet alive that sometimes it is as if we were born on completely different planets. Big changes in society can happen even in just five years. All it takes is one natural disaster like a hurricane or an earthquake to completely remake the social fabric of a community. The United States took a census in 2000, but just three years later, Hurricane Katrina rearranged huge sections of the population in the South, in some cases permanently.
By my count, there are multiple countries that have not had a census in more than 20 years. Afghanistan is in the midst of rectifying this right now. Other countries include: Lebanon, Myanmar, Uzbekistan, Angola, Eritrea, and Somalia.
Another problem is that people are not all equally willing to participate in the census. In some countries – as it was 2,000 years ago – the census is mandatory. But in the United States, you don’t have to answer your door when the census taker comes. I had a friend who took a job as a census taker during the 2010 census, and he was regularly shouted at and berated. If everyone doesn’t participate, that can lead to an uneven view of a community. Those who do provide more detailed information may skew toward being better educated or earning more money. (Unless, of course, they earn so much money that they can afford a razor-wire-topped fence and 24-hour security guards.)
Lastly, the census counts many, many things, but it can’t possibly count everything. That’s where more highly specialized surveys come in. I’ll talk about those in my next post.
Image via census.gov