Associated Press poll on reading

The Associated Press just released a new poll (Aug '07) on American adults' reading habits. Overall, I think the existing surveys I've cited before are higher quality. This one digs deeper into demographic data, but really the National Endowment for the Arts (Reading at Risk) report does better at that too.

One of these polls is not like the others

Some of the surveys I've previously cited focused on literature (fiction, short stories, and poetry). The new AP poll includes nonfiction. The previous surveys I've cited were based on a far larger sample size and used more rigorous methods:

• The Census long form was sent to 17 million households
  
• The National Endowment for the Arts survey interviewed 17,135 people using a mix of in-person and phone interviews
  
• The new AP poll only did 1,003 random-number-dialing phone interviews, with a ±3% margin of sampling error.

Some of the new poll's results are in line with the earlier surveys. Others don't seem to fit the pattern. Still others seem a bit over-interpreted in the AP article. (See How much do we read? and Who reads genre fiction?)

How much do we read?

The new AP poll finds that

One in four adults [27%] read no books at all in the past year.

So 73% of the population did read a book? That's much higher than any of the other surveys have reported. The 2000 Census found that 43% of adults had read a book that year. The 2002 National Endowment for the Arts' Reading at Risk survey found that 57% of adults read a book that year (47% read fiction). (See table.)

How many books?

The AP looked at how many books the typical person read. I'm not sure their numbers mean much--comparing them to the Gallup polls, it's pretty clear the number wanders around from year to year, and from poll to poll. And frankly, it's a bad type of question to ask in a survey. Could you remember exactly how many books you read last year? I couldn't. Only those who read very little and read a book that made an impression on them are likely to answer accurately.

So having said the numbers are crap, here's what the AP says:

The typical [median] person claimed to have read four books in the last year -- half read more and half read fewer. Excluding those who hadn't read any, the usual number read was seven.

These values are close to the Gallup Poll's 2005 findings on how many books people had started in the last year. However, Gallup's results vary from year to year: six in 1990, ten in 1999, five in 2005. Now AP says the typical number is four. I'm not sure any of this is trustworthy.

Gender and country

The AP found that those who read books typically read nine books for women and five for men. Eh... I think this style of graph works a bit better because, failing exact numbers, you can do qualitative comparisons:

Average number of books read or started, 2004. Gallup.
(Note that the Gallup graph uses averages, which will be higher than the medians reported in the AP article. Unfortunately they use two different kinds of statistics, which can't really be compared.)

Who reads?

The AP poll says

Who are the 27 percent of people... [who] hadn't read a single book this year? Nearly a third of men and a quarter of women fit that category.

The AP poll also found that the heaviest readers are those with college degrees, and those over age 50. Neither of those findings is controversial. (See my post with statistics on education and reading.)

I think the new AP poll stretches their data too far in defining regional reading patterns:

People from the West and Midwest are more likely to have read at least one book in the past year. Southerners who do read, however, tend to read more books, mostly religious books and romance novels, than people from other regions.

The National Endowment for the Arts found that the order went West, Northeast, Midwest, then South. Really, it looks like the top 3 regions are within the poll's margin of error; the best statement may be that the West, Midwest, and Northeast have similar numbers of book readers, while the South lags behind in number of readers. The RWA survey agrees that those who read in the South read a lot of romance.

A few last tidbits from the AP poll:

Who reads

• Whites read more than blacks and Hispanics
  
• Those who said they never attend religious services read nearly twice as many as those who attend frequently
  
• Democrats and liberals typically read... slightly more books than Republicans and conservatives

What do we read

• The Bible and religious works were read by two-thirds in the survey
  
• Popular fiction, histories, biographies and mysteries were all cited by about half
  
• One in five read romance novels.
  (This agrees with the RWA survey.)
  
• Every other genre -- including politics, poetry and classical literature -- were [sic] named by fewer than five percent of readers.
  
• More women than men read every major category of books except for history and biography.

Sources

• Alan Fram, Associated Press Writer. August 21, 2007. "One in Four Read No Books Last Year". Chicago Tribune.
  
• National Endowment for the Arts, 2004. "Reading at Risk", 2002 Survey on Public Participation in the Arts. ("The 2002 SPPA asked respondents if, during the past 12 months, they had read any novels or short stories, plays, or poetry... including popular genres such as mysteries, as well as contemporary and classic literary fiction. No distinctions were drawn on the quality of literary works.")
  
• Gallup Poll Tuesday Briefing, Jan 2005, p 77. "During The Past Year, About How Many Books, Either Hardcover Or Paperback, Did You Read Either All Or Part Of The Way Through?"
  
• U.S. Bureau of Census, 2002. "Statistical Abstract of the U.S." National Data Book, Table 1223, p 753.