Know Your Data

Back in the old days, when I was just a kid pretending to be a demographer (the fire ant Census was particularly difficult, by the way), the Census Bureau issued its data in an old-fashioned but remarkably useful medium known as paper. This meant page after page of eyesight-challenging numbers. At the time it seemed like a huge amount of data. How naive we were!

But paper, youngsters, is a more limiting medium than cyberspace. The data released by the Census Bureau had to fit into the paper volumes they published and mailed out to eager data miners. These limitations still shape much of what the Census Bureau publishes on the Internet, since users often want to be able to see trends over time, and maintaining the same data sets (and presentation formats) makes it easier to do this.

But some data, such as population statistics, is now available in basically raw form. This means that rather than five-year age groups, you can get monthly population estimates by single year of age. This provides data users with the opportunity to create their own age groups, such as 18-to-49 year olds, but it also means that if you want five-year age groups, you will have to add them up yourself. (Or—shameless plug alert—get them from New Strategist Publications. Motto: “Be Glad We Can Add.”)

As the amount and level of data detail swells, it also becomes a bit trickier to navigate through all the different data sets and sources. You might not know, for example, that a particular data set on the Census website was discontinued years ago and is now obsolete. (Household projections.) Or you might be confused about which set of population estimates to use since they are now available monthly.(Use the July 1 estimates.)

As yet another service to all eight of our loyal readers, we are instituting a new feature: Know Your Data. This series will answer your burning questions (ACS or CPS??) as well as questions you never thought to ask (Who created the original poverty threshold?). Feel free to email questions you would like us to address. And stay tuned for our first fact-filled offering in the series, appearing on this blog just as soon as the boss stops looking over this way.

Bet You Didn't Know

Percentage of U.S. households with no
adults who speak fluent English: 3

Source: Census Bureau

TV's "Coveted" Demographic

For the life of me, I will never understand why television advertisers are so fixated on the 18-to-49 demographic. This morning, as I perused various online news sources, I came across an article on MSNBC headlined Baby boomers upset TV isn't all about them.

Although the headline suggests this is just more whining from spoiled Boomers, it is clear from the article that the television industry (and I would argue the entertainment industry in general) suffers from myopia when it comes to "older" viewers. Advertisers are charged higher rates for television programming that attracts viewers under age 50, and in particular those under age 40. This means that when Boomers get home from a hard day at the office, turning on the television may not be the relaxing experience they desire. Rather, it may be an exercise in frustration as they wade through program after program aimed at a much younger audience.

The theory seems to be that capturing young adults as customers will create a cadre of loyal buyers for life. But this is ridiculous in the 21st century when "brand loyalty" is an oxymoron. Rather, marketers need to look at who has the inclination and the bucks to buy their products. More often than not, this means the 40-plus age group. Not that I really expect the entertainment industry to suddenly change its focus. But eventually, advertisers may change their minds for them as they become more conscious of the need to appeal to older customers.

Smack Dab in the Middle

If you draw two lines—one running North-South, the other East-West—so that each divides the population in half, the point of intersection is known as the "median center," or put more plainly, Smack Dab in the Middle. Located in what East and West Coasters like to call Fly-Over Country, the precise location of Smack Dab has been shifting since the nation's early days when most folks lived east of the Mississippi River. For a long time it was located in Ohio, but since 1950 it has moved steadily southwest. As of the last Census in 2000, Smack Dab had relocated to Daviess County, Indiana.

Curious about the real Smack Dab, I googled Daviess County, Indiana. It might be the center of the U.S. in a very strict sense, but it seems it could hardly be further from the center of U.S. culture. This is Amish Country. Its Visitor's Bureau promotes a tour of an Amish Village and a visit to the Black Buggy Amish Restaurant and Bakery. You can shop at country stores and take home a handsewn quilt.

While this gives you a quite charming picture of the of nation's median center, if you are seeking more prosaic enlightenment, you will want to turn to the Census Bureau, a treasure trove (or impenetrable jungle, depending on your viewpoint) of information. There is a QuickFacts site, for example, where you get a U.S. map, ripe for the clicking. Click on Indiana and up pops an informative data table. If you need information on a more specific location within Indiana, there are buttons at the top of the page that allow you to select a specific city or county. The Daviess County data table doesn't tell you that it is “Amish Country,” but it does reveal that there are 30,466 inhabitants residing in 10,894 households. More than two-thirds own their homes and the county's median household income is $35,967. The table includes statistics for all of Indiana for comparison.

Or you could just turn to Wikipedia, where you would discover that in April 2006 Daviess County switched time-zone allegiance. But then, it does seem more fitting that the center of the U.S. population would set its clocks to Central Time rather than Eastern Time.

Generation Gap

Median age of the U.S. population by race and Hispanic origin:
(the median is the age at which half are older and half are younger)

White non-Hispanic 40.4
Asian 35.1
Black 31.3
Hispanic 27.2

Source: Bureau of the Census, American Community Survey

Rankings: Born in State

Take a guess: Among the 50 states, which one ranks first in the proportion of (native-born) residents who were born in the state? The surprising answer is New York. According to the Census Bureau's 2005 American Community Survey rankings, 82.3 percent of New York's residents were born in-state. Other states with populations staying close to home are Louisiana (82.0 percent), Michigan (80.5 percent), Pennsylvania (80.1 percent), and Ohio (77.9 percent).

The national average is 67.5 percent. In other words, two-thirds of Americans born in the United States live in the state in which they were born. The state with the fewest residents born in-state is Nevada, with only 26 percent of residents born there. Also close to the bottom are Florida (40.9 percent), Arizona (41.1 percent), Alaska (42.7 percent), and Wyoming (43.1 percent).

Bet You Didn't Know

Percentage of people aged 18 to 65 who have not been to a doctor or other health care professional in the past year, by health insurance status:

With health insurance: 15 percent
Without health insurance: 46 percent

Source: National Center for Health Statistics, Summary Health Statistics for U.S. Adults: National Health Interview Survey, 2005

Feeling Worse

Why don't Americans feel as good as they once did? The percentage of adults aged 18 or older who say they are in "excellent" or "very good" health fell from 59 to 54 percent between 1995 and 2005, according to the government's Behavioral Risk Factor Surveillance System.

The aging of the population explains some of the decline, since older Americans are less likely to report being in tip-top shape. But an examination of the data by age group reveals young adults to be the ones with the biggest decline in health status. The percentage of 25-to-34-year-olds who say they are in excellent or very good health fell by nearly 8 percentage points between 1995 and 2005—from 70 to 62 percent. In contrast, among people aged 65 or older, the percentage reporting excellent or very good health fell by less than 1 percentage point, declining from 37.2 to 36.5 percent during those years.

Every demographic segment was less likely to report being in excellent or very good health in 2005 than in 1995, but some recorded bigger declines than others. By race and Hispanic origin, Hispanics experienced the sharpest drop in health status—down an eyebrow raising 17 percentage points. By income group, those in the middle saw the biggest decline in health status. The percentage reporting excellent or very good health fell by 11 to 12 percentage points for those with household incomes ranging from $15,000 to $49,999. By education, high school graduates and those with only some college experienced a larger decline in health status than those without a high school diploma or a college degree.

One possible explanation for our worsening health status is the decline in health insurance coverage. The groups experiencing the biggest losses in health insurance coverage are also the groups with the biggest declines in health status. The percentage of people aged 25 to 34 without health insurance grew from 23 to 26 percent between 1995 and 2005—the biggest increase among age groups, according to the Census Bureau. This is also the age group reporting the biggest decline in health status. The only age group with universal health insurance coverage—people aged 65 or older—experienced the smallest change in health status during those years. Likewise Hispanics, the group most likely to be without health insurance, saw the biggest decline in health status. Middle-income groups—with incomes too high to qualify for Medicaid (the government's health insurance program for the poor) and too poor to afford employer-provided or private health insurance—also experienced the biggest declines in self-reported health status.

If the lack of health insurance is behind Americans' growing unease about their health, then we're likely to feel even worse in the years ahead as health insurance costs climb and coverage falls.

Bet You Didn't Know

Percentage of children in 4th through 8th grade who take part in organized activities (not including after-school programs) by family income level:

incomes below poverty level: 30 percent
incomes 200% or more above poverty level: 66 percent

Source: America's Children in Brief: Key National Indicators of Well-Being, 2006

Married Couples a Minority

The New York Times reports today that married couples are now a minority of households. The Times' analysis of the Census Bureau's 2005 American Community Survey (ACS) shows that married couples account for 49.7 percent of households, down from 51.7 percent in 2000. The Census Bureau's other trend-tracking survey still puts married couples in the majority, however. The 2006 Current Population Survey (CPS), taken in March, estimates that married couples accounted for 50.9 percent of households. The difference in the two figures is most likely due to the inherent variation in estimates based on population samples.

It does not matter whether married couples are a slight majority or a bare minority of households. The fact is, living arrangements in the U.S. are very different from what they once were and will change even more as the baby-boom generation ages into the empty-nest years and beyond. Here's how living arrangements rank today, from most to least popular according to both surveys:

1. Married couples without children under age 18 living at home (most are empty-nesters)
ACS 28.0%
CPS 27.1%

2. People living alone (most are women and many are widowed)
ACS 27.1%
CPS 26.6%

3. Married couples with children under age 18 at home
ACS 21.7%
CPS 23.8%

4. Female-headed families (about two-thirds include children)
ACS 12.6%
CPS 12.3%

5. People living with nonrelatives (many are cohabiting couples)
ACS 6.0%
CPS 5.7%

6. Male-headed families (only half include children)
ACS 4.6%
CPS 4.5%

As boomers age and many become widowed, living alone is destined to become the most common lifestyle in the U.S.

Bet You Didn't Know

Percentage of homeowners with no public transportation available in their area: 50 percent.

Source: Bureau of the Census, 2005 American Housing Survey

How Many Illegals are in the U.S.?

According to a government report, 10.5 million unauthorized immigrants were living in the U.S. in 2005, up from 8.5 million in 2000—a 24 percent increase. Fifty-seven percent are from Mexico, followed by much smaller numbers from El Salvador (4 percent), Guatemala (4 percent), India (3 percent), China (2 percent), and other countries.

These figures come from the Department of Homeland Security. They were calculated using the "residual method" of population estimation. Researchers in the Office of Immigration Statistics subtracted the number of legal foreign-born residents (based on administrative data at the Department of Homeland Security) from the total number of foreign-born (based on estimates from the Census Bureau's American Community Survey), and the difference—or residual—is the number of illegals in the U.S.

The number of unauthorized immigrants is growing by 408,000 per year, according to the report.

Not surprisingly, California is home to the largest number of unauthorized immigrants—2.8 million or 26 percent of the total in 2005. Other states in the top five are Texas (1.4 million), Florida (850,000), New York (560,000), and Illinois (520,000). But the number of illegals is growing the fastest in Georgia, where it has more than doubled between 2000 and 2005—climbing from 220,000 to 470,000 between 2000 and 2005.

Bet You Didn't Know

Percent moving between 2004 and 2005...
All ages: 14 percent
Aged 20 to 24: 30 percent
Aged 62 to 64: 6 percent

Percent moving out of state...
All ages: 19 percent
Aged 20 to 24: 16 percent
Aged 62 to 64: 34 percent

Source: Census Bureau, Geographic Mobility: 2004 to 2005

If Cell Phones Could Talk

Eighty-two percent of Americans say they have been annoyed by the cell phone use of others in public places, according to a survey by the Pew Internet & American Life Project. But for demographers and others who love to track the trends, cell phones are more than an annoyance. They are an astonishing technological marvel that has revealed one of our deepest needs.

Before the cell phone, who knew Homo sapiens wanted to talk so much? Our prehistoric ancestors might have known, since they were the last to be within shouting distance of everyone they knew almost all the time. The African savannahs might have been humming with conversation as ancient humans discussed the latest hunt and gather. Then, as our species populated and migrated to distant lands without any means of communication with those left behind, we were forced to repress our need to stay in touch. Perhaps out of this melancholy longing art was born. We told stories, wrote ballads and poems, and drew pictures to mourn our loss of communication with loved ones far away. Finally, we invented the snail mail system to maintain some vestige of contact, however inadequate.

It has been thousands of years since we could strike up a conversation at will with any friend or family member. Only in the past ten years, as cell phones proliferated and the cost of using them dropped to pennies in our pocket, have we rediscovered our addiction to continuous communication with those we care about.

The cell phone's evolution over the past decade, from clunky and expensive accessory to sleek and cheap necessity, has revealed talk to be one of our most powerful urges. What we say is far less interesting than the act of saying it, all the time: "Where are you?" "What are you doing?" "Where should we meet?" "How about this for a plan?"

The cell phone's overarching importance is revealed not just by anecdotes, but in the spending statistics collected by the Bureau of Labor Statistics in its annual Consumer Expenditure Survey. In the past decade, the cell phone's significance has grown more than any other household item. In 1997, only 10 percent of households spent on cell phone service. In that year, in a ranking of items on which the average household spends the most, cell phone service was lost in the crowd at 112th place. In 2004 (the latest data available), 46 percent of households spent on cell phone service, boosting it to a lofty 28th place. Cell phone service now ranks higher in our spending priorities than prescription drugs and is not far below cable TV and alcoholic beverages.

Interestingly, while 82 percent of the public has gotten annoyed with someone else's cell phone use, only 8 percent of cell phone users say they have annoyed others with their cell phone calls. They must have been too busy talking to notice.

Does Religion Influence Sexual Behavior?

Does religious belief influence sexual attitudes and behavior? Most would call the question a no brainer--of course it does! But human nature being what it is, religious belief has less influence on behavior than on attitudes.

This becomes clear in an examination of the National Survey of Family Growth, a survey taken every few years by the National Center for Health Statistics. The NSFG probes the attitudes and behavior of a representative sample of Americans aged 15 to 44 toward sexuality and reproduction. An examination of the results suggests that religious belief affects the talk much more than the walk.

Religious belief shapes attitudes toward premarital sex:
The percentage of women who think it is all right for unmarried 18-year-olds to have sexual relations if they have strong affection for one another ranges from a high of 74 percent among women with no religion to a low of 29 percent for fundamentalist Protestant women. Among Catholic women, the 54 percent majority thinks premarital sex is OK.

But it delays first sexual intercourse by only a few months:
Among women aged 15 to 44 with no religion, average age at first sexual intercourse was 16.4 years. Fundamentalist Protestants held off for several more months, with an average age of 16.9 years at first sexual intercourse. Catholics just said no for nearly one year longer, with an average age at first sexual intercourse of 17.7 years.

And it has little impact on "saving oneself" for marriage:
Among women with no religion, only 12 percent waited until marriage before having sex. The proportion was a slightly higher 15 percent among Catholics and 17 percent among fundamentalist Protestants.

Religious belief drives attitudes toward out-of-wedlock childbearing:
Among women with no religion, fully 86 percent think it is OK for an unmarried woman to have a child. A 49 percent minority of fundamentalist Protestants agree. (Among fundamentalist Protestant men, the figure is an even smaller 37 percent.). A surprisingly large 72 percent of Catholic women think out-of-wedlock childbearing is OK.

But it has almost no effect on out-of-wedlock childbearing itself:
Among women aged 15 to 44 with children, the percentage who have ever had a child out-of-wedlock is similar regardless of religion. Among women with no religion, 49 percent have had a child out of wedlock. The proportion is 47 percent among fundamentalist Protestants and a slightly smaller 40 percent among Catholics.

Bet You Didn't Know

Households with two or more earners, by household income quintile:

lowest quintile (with incomes below $19,178): 5 percent
highest quintile (with incomes above $91,705): 76 percent

Source: Census Bureau, 2005 Annual Social and Economic Supplement

Comparing Crime

Recently, the FBI released preliminary numbers from its 2005 Uniform Crime Reporting Program. According to the findings, violent crime in the U.S. rose 2.5 percent between 2004 and 2005. The complete report will be released later this year, entitled Crime in the United States, with tables of data on crimes reported to nearly 17,000 law enforcement agencies across the nation. But don't plan on using the report to determine how your local area ranks in the crime statistics. Those who want to compare one area with another are on their own.

The FBI makes no secret of the fact that they do not want users to compare crime rates. They even write extensively about the difficulty of doing so on their web site:

"Each year when Crime in the United States is published, many entities—news media, tourism agencies, and other groups with an interest in crime in our Nation—use reported figures to compile rankings of cities and counties. These rankings, however, are merely a quick choice made by the data user; they provide no insight into the many variables that mold the crime in a particular town, city, county, state, or region. Consequently, these rankings lead to simplistic and/or incomplete analyses that often create misleading perceptions adversely affecting cities and counties, along with their residents"

But you have to wonder whether this high-mindedness is more political than methodological. After all, if the FBI provided rankings of states and local areas by crime rate, then those 17,000 law enforcement agencies might not be so cooperative about sending their data to the FBI. And more than 400 congressional representatives, 100 senators, and countless local politicians would be very unhappy if their area appeared at the top of a high-crime list.

Unless an intrepid reporter burrows into the numbers by downloading, multiplying, dividing, and sorting, the average American is left pretty much in the dark about relative crime rates in the U.S. Everyone knows the central cities of metropolitan areas have high crime rates, but what about college towns, vacation spots, and retirement areas? The federal government makes it hard to find out.

Recently, crime seems to have surged in my town, Beaufort, South Carolina—a military base and resort area on the coast. Letters in the local newspaper have expressed shock at the crime, wondering what is happening to our once safe community. But how safe is Beaufort, relatively speaking? I wanted to find out. Because the 2005 data are not yet available, I went to the FBI's 2004 report, finding a table listing crime rates by region and state. I was not surprised to see the South having the highest rate of violent crime—this is a well documented fact. By region, the violent crime rate per 100,000 population looks like this:

U.S. average 465.5
South 540.6
West 480.7
Midwest 391.1
Northeast 390.7

Next, I spent about 15 minutes sorting the 50 states by their violent crime rate. To my surprise, supposedly sleepy South Carolina came out on top, with 784.2 violent crimes per 100,000 population in 2004. The violent crime rate in South Carolina exceeds the rate in every other state, including those with high-crime reputations such as Florida (711.3), Nevada (615.9), Texas (540.5), and New York (441.6).

But what about Beaufort? Another table has the data, listing each city's population and the number of violent crimes reported in 2004. The rate has to be calculated, then the list sorted. Lo and behold, Beaufort's violent crime rate of 1,660.3 places it 7th out of 33 cities in the state with populations of 10,000 or more. Beaufort, in fact, has a higher crime rate than the central cities of the state's largest metropolitan areas—Columbia and Charleston. Beaufort has a lower crime rate, however, than the state's other resort area, Myrtle Beach.

While it is understandable that local businesses and politicians may not want these facts to be released, keeping residents in the dark about their chances of experiencing violent crime is a disservice to the community. That kind of naiveté not only increases risk taking, but also adds more victims to the crime rolls.

Problems and Who Has Them

Every now and then someone does a study that tells it like it is. This is the case with "Troubles in America: A Study of Negative Life Events Across Time and Sub-groups," by Tom Smith, Director of the General Social Survey of the University of Chicago's National Opinion Research Center. The study, prepared for the Russell Sage Foundation, is also mentioned in the post below and could be the subject of dozens of posts examining its many findings.

In both the 1991 and 2004 GSS surveys, a representative sample of Americans was asked whether they had experienced any of 66 problems during the past year. The incidence of problems alone is worth pondering. Here is a look at the some of the troubles and the percentage of Americans experiencing them in the past 12 months, based on the 2004 survey:

PERCENTAGE EXPERIENCING IN PAST YEAR
Sick enough to go to a doctor: 56.2
Death of a close friend: 22.0
Lacking health insurance coverage: 17.9
Major home repairs: 15.4
Being unemployed for as long as a month: 14.5
A major worsening of one's financial condition: 13.1
Not having a car for at least one month: 8.3
Undergoing counseling for emotional problems: 7.4
A cut in pay: 6.8
Having serious trouble with a child: 6.3
Being discriminated against: 4.7
Death of a parent: 3.1
Getting divorced: 2.7
Home destroyed or damaged by fire, flood, etc.: 2.0

The Decade of Retreat

If the 1990s is known as the Decade of Affluence because household incomes reached a peak, then these years may well be called the Decade of Retreat—especially for the American worker. Hardly a day goes by without more evidence of workers falling behind.

The Current Population Survey data, released yesterday, show a signficant decline in the earnings of men and women who work full-time (see post below). Wages and salaries now account for a smaller share of the gross domestic product than at any time since 1947, according to an analysis by the New York Times. Not only are wages falling, but the percentage of Americans with employment-based health insurance is slipping, and those with insurance must pay more for it. In fact, the value of employee benefits is failing to keep pace with inflation, according to the Times.

Americans are feeling the pain. According to a survey by the Pew Research Center, 59 percent of the public thinks today's workers must work harder to earn a decent living. Fifty-six percent say employers are less loyal to workers. Fifty-one percent think retirement benefits are not as good as they used to be. A comparison of the 2006 Pew results with surveys taken in 1989 and 1997 shows job satisfaction dropping sharply among workers aged 50 or older and growing discontent with employee benefits.

The quality of work life may not be alone in retreat. Tom Smith, director of the General Social Survey (GSS) at the University of Chicago's National Opinion Research Center, concludes in a study that "Americans on average are worse-off now than in the past." He came to this conclusion after comparing 2004 with 1991 GSS results. In both years, respondents were asked to identify the frequency and severity with which they encountered 66 different problems ranging from the death of a spouse to the loss of a job and the inability to pay for health insurance. The percentage of Americans experiencing at least one problem increased over the years from 89.1 to 91.5 percent. The average number of problems experienced grew from 4.03 to 4.34. Smith found significant increases in problems such as being unemployed, being pressured to pay bills, being unable to afford food, lacking medical insurance, and being hospitalized.

What factors might explain an erosion in the quality of life in the United States? One might be the aging of the population, which leads to more health problems. Another might be technological change and the consequent globalization of the workforce. The Pew survey shows how ambivalent Americans feel about these changes. Sixty-nine percent think email and other modern communication technologies have helped workers. But an even larger 77 percent think outsourcing—which is one consequence of the communications revolution—has hurt American workers.

Smith's study sheds some light on why things may be turning south, examining the demographic segments with the most and least troubles. The biggest change in the distribution of troubles was by education. In 1991, the lowest-educated group experienced 1.33 problems for every 1 problem experienced by the best-educated group. In 2004, the ratio had grown to 1.79 to 1. The growing gap between the haves and the have-nots in an increasingly competitive world may explain the Decade of Retreat.

Here Are the 2005 Numbers

At 10:00 this morning, the Census Bureau unveiled the much anticipated snapshot of our economic wellbeing, the results of the Current Population Survey's Annual Social and Economic Supplement. Taken in March of each year, the supplement asks a large, nationally representative sample of Americans about their economic status in the previous year. Here are the numbers unveiled by the Census Bureau this morning.

HOUSEHOLD INCOME UP
Median household income rose 1.1 percent between 2004 and 2005, to $46,326. This is the first significant increase in median household income since 1999, according to the Census Bureau. But the 2005 figure is still well below the level of 2000, and the gains were limited to a few select demographic segments. By age, only householders aged 65 or older saw their median household income rise significantly between 2004 and 2005—up by 2.8 percent. Behind the gain is higher labor participation among the elderly. By income, changes were not significant at the 90 percent confidence interval, but it is interesting to note that all but the highest income quintile lost ground.

MEN'S EARNINGS DOWN
Among men working full-time, median earnings fell 1.8 percent between 2004 and 2005, to $41,386 after adjusting for inflation.

WOMEN'S EARNINGS DOWN
Among women working full-time, median earnings fell 1.3 percent between 2004 and 2005, to $31,858 after adjusting for inflation.

HEALTH INSURANCE DOWN
The percentage of people without health insurance climbed from 15.6 to 15.9 percent between 2004 and 2005. The number of uninsured grew to 47 million. The percentage of the population covered by employment-based health insurance fell from 59.8 to 59.5 percent.

POVERTY STABLE
The percentage of people living in poverty held steady at 12.6 percent in 2005, not significantly different from the 12.7 percent rate of 2004.