Friday, May 30, 2008

Maternal employment

Today, most mothers are employed, and the majority of employed mothers are working full time. This is especially true of mothers of older children, but even among mothers of infants, half (51.9%) are employed and by far most of these mothers are employed full-time.

These high levels of maternal employment reflect a substantial social change since the 1960s and 70s. Yet, maternal employment has not grown substantially since the late 1990s, despite the passing of legislation (e.g. the Family Medical Leave Act, 1993) that should have improved women’s ability to combine paid work with child rearing.

Trends in maternal employment spark intense controversy between those who argue that women should work, either for greater self-fulfillment or because it is very difficult for most families to make ends meet without the mom working. Others, take offense at the notion that caring for children is not noble or worthy, sometimes demeaning mothers who do decide to take one paid employment. I can understand this perspective, since we often devalue care work.

A consequence of this controversy is that we often
scrutinize trends to find evidence that women are starting to forgo career for family (or vice versa). Either out of economic necessity or philosophical differences, diversity is here to stay.

Thursday, May 29, 2008

Aging Population

The population of the United States is aging remarkably slowly. Today 12.4 percent is over age 65 and this actually represents a decrease from 12.6 percent in 1990. In contrast, Japan is experiencing a fairly rapid aging of its population. In 2000, 17.3 percent of the population was age 65+, up from 12.0 percent 10 years earlier. We can expect that the United States will age a little bit faster as the baby boomers (born 1945-1964) reach their golden years, but nothing like Japan.

Wednesday, May 28, 2008

Demographic data

According to a recent report by the National Center for Health Statistics, almost 1 out of 6 United States residents lives in a household without a landline. This has important implications for the way basic demographic data are collected, particularly for young adults (age 18-24). In an attempt to maintain data quality, some survey organizations are starting to add cell phone numbers to their call lists.

Not including cell-phone only households is not a problem if these households are similar to households with landlines, but this same study found for instance that binge drinking (having 5 or more drinks in an episode) was twice as common among the cell-phone only households.

Click on this picture to actually see it:

Tuesday, May 27, 2008

Variation in Education

The American Community Survey is a good source of up-to-date information on population characteristics. For example, they have state ranking tables that show facts such as median age, the sex ratio, or percentage of the population with a bachelor’s degree. Washington D.C. has an exceptionally high proportion of the population with a bachelor’s degree. It also has the highest rate of child poverty. That is, Washington D.C. has an exceptionally high level of inequality, but this is not really a fair comparison and we would likely find similarly high levels of inequality in New York city or many other large urban areas.

"States" ranked by percentage of Population with a Bachelor's Degree
West Virginia 17
Arkansas 18
Mississippi 19
Kentucky 20
Louisiana 20
Nevada 21
Alabama 21
Indiana 22
Tennessee 22
Oklahoma 22
South Carolina 23
Wyoming 23
Ohio 23
Idaho 23
Iowa 24
Missouri 24
Michigan 25
Texas 25
North Carolina 25
South Dakota 25
Wisconsin 25
Florida 25
New Mexico 25
Pennsylvania 25
Arizona 26
North Dakota 26
Maine 26
Georgia 27
Alaska 27
Nebraska 27
Delaware 27
Montana 27
Oregon 28
Kansas 29
Utah 29
Illinois 29
California 29
Rhode Island 30
Hawaii 30
Minnesota 30
Washington 31
New York 31
New Hampshire 32
Vermont 32
Virginia 33
New Jersey 33
Connecticut 34
Colorado 34
Maryland 35
Massachusetts 37
Wash. D.C. 46

Monday, May 26, 2008

Remembering Veterans

Memorial day is an opportunity to remember those who have died in the military service. By the most recent count I can find, 4,081 men and women in military service have died in Iraq and 432 have died in military service in the Afghanistan area since 2001. Knowing how difficult it is to lose these lives helps put into perspective the 11,260,000 men and women who died in military service to the United States in World War II and the 58,209 who died in Vietnam.

Generally speaking more recent generations have borne a much lighter military burden than older generations. According to the 2006 American Community Survey about 20% of the male population in the United States has served some time on active duty in the U.S. Army, Navy, Air Force, Marine Corps, or the Coast Guard.* This proportion is much lower for younger generations than those who were old enough to fight in WW II.

These generational differences will likely shape the debates between Obama and McCain this summer. McCain, who will turn 73 this August, is of a generation where the majority of men served in the military. Obama is 46 and of a generation where military service was much less common.

* The full definition of veteran that the Census bureau uses is as follows: a person 18 years old or over who has served (even for a short time), but is not now serving, on active duty in the U.S. Army, Navy, Air Force, Marine Corps, or the Coast Guard, or who served in the U.S. Merchant Marine during World War II. People who served in the National Guard or military Reserves are classified as veterans only if they were ever called or ordered to active duty, not counting the 4-6 months for initial training or yearly summer camps.

Friday, May 23, 2008

Marriage again

Last Friday I showed a graph depicting the increasing age at marriage. The interesting aspect of this trend is that it means that we are delaying an event that nearly everyone looks forward to. Typically we like to “eat dessert first” so to speak, so this is a bit of a departure from the norm. For a minority of folks, gays outside of Massachusetts and California, perhaps legal constraints account for the delay, but for most this isn’t a good explanation.

Up to a point, delayed marriage is wise. Waiting until age 21 substantially
decreases the chances of divorce.After age 21, delays don't do much to reduce the risk of divorce, but the median age at marriage is over 27 for men and over 25 for women. So, why wait 4-6 years after age 21?

Could it be that we actually value something else more than marriage? Or is it just that we’d like to marry, but the right person hasn’t appeared….yet.

Thursday, May 22, 2008

Unconsidered Patriot

What is patriotism? Is it wearing a lapel pin? Is it supporting our troops? Could it be speaking out against war? In 2004 the General Social Survey asked a sample of United States residents whether people should support their country even if the country is in the wrong. Approximately the same proportion agreed or strongly agreed with this statement (37%) as disagreed or strongly disagreed (40%). The propotion responding affirmatively for those with less than a high school diploma is considerably higher (46%) than for those with a college degree (29%).

It seems that even elementary education teaches that patriotism (in the United States) is about the ideals of democracy and not simply about proudly waving the flag. Why isn't there more widespread disagreement with this statement across all education groups?

Wednesday, May 21, 2008

Diet and Nutrition

A recent report released by the National Center for Health statistics describes the dietary habits of those age 60 and older. The report provides information on overall quality of diets as well as specific aspects. Dietary guidelines suggest that the typical woman should have 6 servings of grains, 3 servings of vegetables, 2 servings of fruit, 2-3* servings of dairy, and 2 servings of protein daily.
*The number of servings depends on your age. Older children and teenagers (ages 9 to 18 years) and adults over the age of 50 need 3 servings daily. Others need 2 servings daily. During pregnancy and lactation, the recommended number of milk group servings is the same as for nonpregnant women.
Overall, the NCHS report indicates that fewer than one in three elderly Americans get sufficient amounts of the 5 food groups and only 17% have a "good" quality diet. The graph below presents some of the details by food group for women. The height of each bar indicates the score for each food group by education. If 100% of women ate sufficient amounts of meat daily, the height of the bar would be 10. The lightest bars are for those with less than a high school degree; the medium green bars are for those with a high school degree but no college and the darkest bars are for those with at least some college.
Deficiencies are greatest for fruits, especially for those with less than a high school degree. The patterns for men look very similar (not shown), except that dairy is tied with fruit for the highest deficiency.
Importantly, eating more often does not lead to a better quality diet. The obese have the greatest deficiency of dairy and fruits.
I guess the good news out of the report is that 72% of elderly adults consume no more than the recommended amount of cholesterol.

Tuesday, May 20, 2008

Kentucky and Oregon

My predictions for West Virginia underestimated by a considerable margin the amount of the vote Clinton would win and I didn’t do nearly as well as some other prognosticators. A variety of factors that I didn’t consider in my estimates worked in her favor. First, West Virginia has a less educated population than any of the states with recent primaries. Obama does better with the college educated crowd than does Clinton. I really don’t think this is the direct reason why my estimates were off, however. I think that the real reason was that West Virginia’s population is older than most other states. Older folks tend to have less education, turn out in large numbers at the polls, and vote for Clinton.

Kentucky’s population more closely resembles that of Ohio or Indiana in terms of age and it falls somewhere between Indiana and West Virginia in terms of the educational attainment of whites age 25+. This leads me to expect that Clinton will not do quite as well in Kentucky as she did in West Virginia. All in all I expect her to get about 56% of the vote in that state – a solid win, but not as much as what is currently projected by
polling data and not enough to change the outcome of the nominations process.

Figuring out Oregon is much more difficult. There haven’t been many recent primaries in the West, few of these primaries were accompanied by exit polls, and clearly the West votes differently than
Appalachia. In short, I lack any useful data. It seems like recent polls indicate that Obama will win 50-55 percent of the vote. I don’t think it will be that close. Obama won 61 % of Wyoming, 68 % of Nebraska and Washington State, 66% of Minnesota and 80 % of Idaho. He did lose in neighboring California, obtaining only 41% of the vote there, but county maps suggest he did better in the northern part of that state. My rough guestimate is that Obama will win at least 64% of the vote in Oregon.

Monday, May 19, 2008


The Bureau of Labor statistics posted this nifty map showing unemployment rates for states on Friday. California, Michigan, Rhode Island, Alaska, and Washington D.C. currently have the highest levels of unemployment. In addition, Florida, Nevada, and Georgia have experienced a percentage point gain in unemployment over the past year.

It's important to note that unemployment has a very narrow definition. To be unemployed a person has to not have been employed in the last week and to have actively looked for work in the last 4 weeks. If we add in people who are discouraged workers (that is, people who want work but haven’t looked because they don’t think jobs are available) and other people who aren’t working as much as they would like, the proportions are somewhat higher.

Meanwhile, Oklahoma has experienced a percentage point drop in unemployment over the past year. I guess it is not much surprise that the nation's breadbasket, as well as many of the oil-producing states, are doing pretty well these days.

Friday, May 16, 2008


A social trend that provokes nearly no one is the increased median age at marriage. Trends in the percent married by the end of their 25th year show remarkable declines in marriage in early adulthood. Fewer and fewer young adults marry in their early twenties. Almost everyone says that they would like to marry someday and most also say that they would be better off if they were married. Yet, most are content to put this off until some time later.

If marriage is such a good thing, why put it off?

Thursday, May 15, 2008


Today's entry demonstrates that the more things change, the more they stay the same. You'd think that, given dramatic changes in the cost of of a gallon of gas, the political party in power, levels of unemployment, racial unrest, divorce, or whatever over the past 26 years there would have been wild fluctuations in levels of happiness. Nope.

Maybe if I choose a different scale for the graph....nope. I have to think that 1984 wasn't a good year, but otherwise things are pretty stable.

Wednesday, May 14, 2008


A recent report published by the National Center of Health Statistics examines sleep patterns in the United States and the association between sleep and other health indicators. Interestingly, most (63%) people get the recommended 7-8 hours of sleep each night, but this leaves a substantial minority (about 38%) getting less sleep than recommended and a small proportion (9%) sleeping 9+ hours a night. Those getting 7-8 hours of sleep have the lowest rates of smoking, inactivity, and obesity.

Of course, it’s hard to say that sleeping too little or too much makes you obese. It might be that being obese makes it harder to sleep. Or it could be that other health conditions both interfere with sleep and exercise. Nonetheless, other research suggests that people who don’t get enough sleep eat more during the day.

Bottom line: A growing number of studies suggest that getting enough sleep is an important part of maintaining good health.

Tuesday, May 13, 2008

West Virginia Primary

I did farily well predicting the North Carolina and Indiana primaries just using exit poll and census data, especially if you ignore my delegate calculations. (Doh!) Let's see how dumb my luck was...
I worked under the assumption that voting tendencies by race and gender would hold steady. And the table above shows that they did. If anything, Clinton grabbed a larger share of the white vote than expected. This was offset by the fact that the voting rates for whites were much lower than for blacks. Obama is clearly drawing many new minority voters to the polls.
West Virginia has an exceptionally large proportion of its white population that never went to college. We can expect these voters to like Clinton over Obama, but lets see if they actually go to the polls now that Clinton's chances of receiving the nomination appear severely dented.
Bottom line: Keeping with what so far as worked pretty well, I predict that Clinton will win West Virginia with 58% of the vote. Currently polls suggest an even larger win. Let's see who's right.

Monday, May 12, 2008

College Squeeze

It seems that the college admissions game continues to get more crazed every year. This is true not only among elite schools, but also public universities. Two issues make getting into a top college much more difficult in recent years. The first is the remarkable growth in postsecondary schooling. Nowadays, most high school graduates continue their educations after receiving their diplomas. Another factor has to do with the number of babies born approximately 18 years ago.

Between 1985 and 1990 the number of babies born in the United States increased from 3.67 million to 4 million. That meant that those applying for college this year were competing with a lot more people than those applying just five years ago. The situation will be even worse next year, but should start to get better in 2010.

Those of you with 7 year olds might start working on buffing up those GPAs and test scores ASAP.

Tomorrow...West Virginia predictions.

Friday, May 9, 2008

Family Life

An interesting trend in family life, one that almost everyone knows about, is the growth in single parent families. So, it isn't much surprise to anyone that many households (60 percent in 2006) with a female householder contain children. It might be slightly more surprising to know that almost half of households with an unmarried male householder have children. Some might see this as evidence for a sudden increase in men taking custody of children in the event of divorce. It is not. No, this is because of an increase in unmarried couples living together with children, sometimes their own children, sometimes children from a previous relationship.

What I find surprising in the above graph is that the household type that is least likely to have children is a married couple household. You might think then that most children are raise in single parent families. This would be wrong...By far, most children live in married couple households. This is possible because most households, with children or not, are maintained by a married couple.

Thursday, May 8, 2008

Trends in Religious Fundamentalism

Reading various online sources, I see that many have impressions about whether religious fundamentalism or religious liberalism is increasing or on the decline. The General Social Survey, conducted by the University of Chicago’s National Opinion Research Center, has asked a representative sample of United States residents about their religious beliefs in every interview since 1972. In the early 1970s most responded that they had moderate beliefs and less than 20 percent claimed to hold liberal beliefs. From the early 1970s through the mid 1980s religious fundamentalism (or at least the proportion claiming to be fundamentalist) was on the rise and there was a substantial decline in moderates. Since that time, the proportion claiming to be liberal is increasing.

I'm not sure whether these trends actually represent real declines in the proportion holding fundamentalist beliefs (e.g. that the bible is the literal word of the creator) or if it reflects growing distaste for the "fundamentalist" label as it is now associated with religious extremism and intolerance. It's hardly scientific evidence, but I found identifying a blog that has a positive spin on "religious fundamentalism" to be difficult, while finding one that is religiously conservative is easy. I think the increase in the proportion reporting that they are liberal (despite the fact that this term is not especially popular) reflects real changes in beliefs.

Wednesday, May 7, 2008

Demography of Aging and Health

Given the aging of the United States (and for that matter, the world) population, health is a growing industry. As explained by UnderstandingSociety there is a similarity between sociology (the home discipline of demography) and epidemiology. One place where these two disciplines actually intersect is in the study of social influences on health and disease. A persistent finding in the health literature is that education has a strong influence on health and mortality. Age-adjusted death rates for both men and women age 25-64 are much lower for those with higher levels of education, as illustrated by these results published in a report by the National Centers for Disease Statistics.

These differences can not be explained by differences in income and is observed across many different countries.

Tuesday, May 6, 2008

Unemployment Trends

According to a report released last week by the Bureau of Labor Statistics the unemployment rate in April was 5.0%, up from 4.5 a year ago, but little changed from last month’s rate of 5.1%. Generally these findings were better news than what was expected, but it’s still the case that employment in construction, manufacturing, and retail trade decreased. This is a very short-term view of labor force change, which might be helpful in determining how long and deep our current recession will be, but monthly shifts can also be over-interpreted.

Taking a longer-term view, we can clearly see that among those age 25 and older, unemployment rates are much higher for those with less education. April 2006 to April 2008 unemployment rates grew 13% for those with less than a high school degree, but actually declined slightly (4.5%) for those with a Bachelor’s degree. Of course, you get different trends if you compare across different time periods, yet a consistency across time is that unemployment rates are less volatile for the more educated.

Tune back tomorrow for a post on Health and Aging.

Monday, May 5, 2008

Primary Trends

Many believe that Rev. Wright’s sermons (as they were presented in the media) did substantial damage to Barack Obama’s advantage in North Carolina. Recent polling data supports this interpretation, but polls don’t present a consistent picture and they aren’t the only source of information we have available to analyze this issue. In addition, we have information on the demographic make-up of Indiana and North Carolina. We also have exit poll data.

Exit poll data show that an increasing proportion of white men and women are voting for Clinton, while Clinton’s share of the black vote is declining. Note that Rev. Wright’s sermons were broadcast around March 10 and Obama’s speech on race (responding to the Wright furor) occurred about a week later, March 18. The Ohio Primary was March 4 and Pennsylvania’s was April 22. Judging just from the trend from Ohio to Pennsylvania, it looks like Obama handled the situation well enough not to lose votes to Clinton, at least not during March and April.

Another way of putting this is, if Pennsylvania had the same demographic profile as Maryland, he would have won over 60 percent of the vote there. Based just on the race and gender distributions of Indiana and North Carolina’s populations, as well as voting rates and voting patterns by race and gender, I predict that Obama will win 56% of the vote in North Carolina and Clinton will win 52% of the vote in Indiana. Or, put another way, Obama will collect 114 delegates to Clinton’s 104 (calculations are available on request).

It may happen that, because I have not completely accounted for the recent downward trend in Obama support among white voters, I am underestimating share of the vote that Clinton will win in Indiana on Tuesday. Yet, I don’t think this is as much an issue in North Carolina since there has also been an upward trend in Obama support among African Americans and Obama continues to hold strong among whites who attended college. North Carolina has a fairly well-educated population.

Here are some tables to support my claim above that if Pennsylvania had Maryland's population composition, it would have been a big Obama win.

Tomorrow's post will be about unemployment trends.

Friday, May 2, 2008

Family Life

A long-term trend in family life has been the increase in the labor force participation of mothers. Given these changes in behavior, it isn’t surprising that most people disagree with the statement that “Family life suffers when mothers work full-time.” Yet in recent years, there has been a notable increase in the proportion agreeing with this statement among both men (solid lines) and women (dashed lines). Source: General Social Survey.

Recently many have tried to link concerns about the growing obesity problem with growth in mother's employment. Of course, we all know that correlation isn't causation. Also, I think some people forget that many more two-parent families would be in poverty if mother’s didn’t work.

Monday we'll post a comment on North Carolina demographics and exactly how bad this Rev. Wright affair would have to be to lead to an Obama loss in that state.

Thursday, May 1, 2008

Indiana's Demographics

After the Pennsylvania primary, much was made of Obama’s senior problem. Exit poll data indicate that Obama has recently done better with younger voters, those with a college degree, and African Americans. Clinton has done better with Anglos and (sometimes) Hispanics, older voters, and those with less education. How did Pennsylvania’s demographics disadvantage Obama? And perhaps more importantly as we head into the Indiana Primary, are Indiana’s demographics similarly advantageous to Clinton?

The answer is mixed. Pennsylvania’s age distribution is older than the rest of the country. Indiana is more similar to the United States as a whole. This works in Obama's favor. Indiana’s race-ethnic distribution is much like Pennsylvania’s although there is a slightly higher proportion of African Americans. Given that Obama gets an extremely high percentage of the black vote, even this small difference can work in his favor, just as the slightly older age distribution in Pennsylvania worked in Clinton’s.

Slightly more hopeful news for Clinton can be found in the education distribution. Indiana has an even lower proportion with a bachelor’s degree than Pennsylvania. Her enthusiasm might be slightly dampened, however, when she finds that this is not because Indiana has more people who stopped their formal education at the end of high school. In Pennsylvania, Clinton won 64% of the votes of high school graduates who didn't go to college. Indiana's actually has a slightly smaller prorpotion of people in this group and proportionately more people who started but didn’t complete college. Clinton’s advantage in this group is not great;
she won 51% of this group in Pennsylvania.

Bottom Line: Overall the demographics of Indiana compared to Pennsylvania indicate that the upcoming race should be closer and and Demography is King.

Come back tomorrow for an analysis of trends in attitudes towards mother's employment.