The Good News About Marriage Part 9

Within the individual sources we cite, you'll find a lot of good data tables (tables of census data, for example). At our website, www.goodnewsmarriage.com, we have a crib sheet of websites we found helpful and used for the book.

If I don't have time to look on your website, can you point me to one good resource?

A good place to start is the Inst.i.tute for Family Studies (IFS). Scott Stanley and Brad Wilc.o.x are senior fellows for this relatively new resource that highlights the latest in marriage research. Their blog even includes ready-to-tweet quotes. The IFS website can be accessed at http://ifstudies.org.

After Tally and I first learned the divorce rate wasn't what we thought it was, we set out to discover the truth-with no idea we were initiating a project that would take eight years, thousands of hours, and the help of hundreds of generous people. Although there is no way to list every name, we want everyone to know we deeply appreciate your help. Please forgive us if our tired brains inadvertently left anyone out. Thank you to...

The many renowned demographers, sociologists, and other experts who have been working in this field for years, who advised, informed, and provided us excellent and often customized data through meetings, conference calls, and e-mails. Most especially we thank Dr. Scott Stanley at the University of Denver, Dr. Brad Wilc.o.x at the University of Virginia, Dr. Sam Sturgeon at Demographic Intelligence, Dr. Jeffrey Dew at Utah State University, director Tom Smith at the General Social Survey, and the incredibly skilled and helpful staff of the US Census Bureau ACS and SIPP offices (who, due to policy, preferred not be named). Our thanks also to Dr. Paul Amato at Penn State University, Dr. Dana Rotz at Mathematica Policy Research, Dr. Linda Waite at the University of Chicago, Dr. David Olson at PREPARE/ENRICH, Dr. Steven Beach at the University of Georgia, Dr. Kelly Raley at the University of Texas, Dr. Betsey Stevenson at the University of Michigan, Dr. Annette Mahoney at Bowling Green State University, Dr. Christopher Ellison at the University of Texa.s.san Antonio, Dr. Tim Heaton at Brigham Young University, Dr. Sheela Kennedy at the University of Minnesota, Dr. Bradley Wright at the University of Connecticut, Dr. Francesca Adler-Baeder and Dr. Chelsea Garneau at Auburn University. Our thanks also to Dr. Andrew Cherlin at Johns Hopkins University; though he was on sabbatical and not directly interviewed, we heavily drew upon his work.

Clint Jenkin and Pam Jacob at Barna Group, who worked with us many months to dig out the real answers about churchgoers, George Barna for his leaders.h.i.+p in the research arena over so many years, and David Kinnaman for his leaders.h.i.+p of Barna Group today.

Dr. Chuck Cowan at a.n.a.lytic Focus and Felicia Rogers and the team at Decision a.n.a.lyst for their expert help and support of my own research surveys over the years, including those in this book. Thanks also to Matthew St. John of the Knights of Columbus for providing valuable information from their commissioned Marist Poll.

Other leading experts and a.n.a.lysts in the marriage arena from a policy, research, therapy, and ministry perspective, who helped us in so many ways. We are especially indebted to the team at FamilyLife, particularly Family Needs Survey director Glenn Gritzon, Greg Weaver, Ron Deal, Chuck Eckerson, Bill Eyster, Bob Lepine, and Dennis Rainey, as well as Dr. Peter Larson at Tango Group. Our thanks also to Jim Daly and the leaders at Focus on the Family, especially Glenn Stanton. Many others provided information, counsel, and feedback during the process, including Pat f.a.gan at Family Research Council, Ed Stetzer, Dwayne Ewers and John Wilke at LifeWay Research, Dr. John Trent, Dr. Wendy Walsh, Diane Mannina at Heritage Foundation, as well as Randy Hicks and Jamie Lord with the Georgia Center for Opportunity, who first helped me understand the basics. We are very grateful to Jim Burns at HomeWord for seeing the potential in this information and encouraging us to share it, and also to Ted Lowe at MarriedPeople, Tyler Reagin at Catalyst, and other leaders for giving us the opportunity to do so.

The dozens of pastors, priests, marriage therapists, psychologists, social workers, counselors, coaches, church leaders, and other faith clergy who answered my call, took meetings, and gave me invaluable insight and perspective, especially Andy Stanley, pastor of North Point Community Church, who wrote the foreword and whose encouragement was invaluable for initially launching me into marriage ministry; and the North Point staff team, especially Diane Grant, Erica McCurdy, Debbie Causey, and Tricia Sherriff, and the counseling team that provided such important input.

The amazing team at WaterBrook Multnomah, especially Dave Kopp, who has been encouraging me in this research for years, our wonderful editor Susan Tjaden, and Ken Petersen, Steve Cobb, Carie Freimuth, Lori Addicott, Allison O'Hara, Laura Wright, Karen Sherry, and Julia Wallace.

While the professional help of those above was vital, we could never have completed the course without the unswerving support, love, and encouragement from our most important people.

From Tally: I am indebted to Shaunti for this unbelievable opportunity and for the last eight years; to Bob Hostetler, who originally recommended me to Shaunti; and to my church, small group, book club, and countless friends, family, and loved ones who supported me during this process. And finally, to my husband, Eric, and our four precious kids, who remind me daily about the good news in marriage. Thanks be to G.o.d!

From Shaunti: I am so amazed and grateful for the skill and pa.s.sion of Tally Whitehead, who started as a research a.s.sistant and turned into a true partner in this work. I'm thankful for my tremendous staff, especially my staff director, Linda Crews, and past and present team members Theresa Colquitt, Kathy Dunmon, Debbie Licona, Cathy Kidd, and Karen Newby, who found themselves running the whole show the last year or two as I disappeared under the weight of four book deadlines. Also Naomi Duncan, Julie Fidler, Jenny Reynolds, and the many others I have depended on in so many ways.

I deeply appreciate my parents, d.i.c.k and Judy Reidinger, as well as Calvin Edwards, Lisa and Eric Rice, my amazing prayer team, and so many other friends for their constant encouragement and practical help.

Most important, none of this would have happened were it not for my wonderful husband, Jeff, and our children, who not only carried me every step of the way, but who showed me the love of G.o.d every day. Ultimately it is His love and guidance that kept pulling me and Tally forward and brought this book into existence, and I thank Him most of all.

1. Ann Patchett, "Ann Patchett Tells Everything She Knows About Love," Reader's Digest, November 2013, www.rd.com/true-stories/love/ann-patchett-tells-everything-she-knows-about-love.

2. Wendy D. Manning, Susan L. Brown, and Krista K. Payne, "Two Decades of Stability and Change in Age at First Union Formation," Working Paper Series WP-13-06, National Center for Family and Marriage Research (Bowling Green State University, October 2013), 22, www.bgsu.edu/content/dam/BGSU/college-of-arts-and-sciences/NCFMR/doc.u.ments/WP/Wp-13-06.pdf.

3. Joyce A. Martin et al., "Births: Final Data for 2011," National Vital Statistics Reports 62, no. 1 (June 2013): 9, table C, www.cdc.gov/nchs/data/nvsr/nvsr62/nvsr62_01.pdf.

4. The words "divorce rate" can mean many different things. In this book, wherever I don't otherwise state the type of divorce rate, I will be referring to the society-wide divorce rate as commonly understood by most people I interview: the percentage of marriages that have ended in divorce. Throughout this book, I refer to it as the "prevalence of divorce," the "current divorce rate," or just the "divorce rate." Wherever I discuss a different type of divorce rate-for example, a projection of new marriages that might end in divorce, the crude divorce rate, various divorce-ratio models, and so on-I will make clear which type of divorce rate I am referencing.

5. Sixty million is a rough middle of the road. According to the American Community Survey 2012, there are approximately 63.7 million current marriages; when respondents were asked if their spouse was present, the number drops to about 57.8 million marriages. See US Census Bureau, 2012 American Community Survey, "s.e.x by Marital Status for the Population 15 Years and Over," http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_12_1YR_B12001&prodType=table.

6. But remember, the expert I first talked to said (see opening of chapter 1), no one actually knows what the real divorce rate is-not just because there are many different types of divorce rates, but because the government discontinued collection of actual divorce certificate data for vital statistics in the late 1990s and has relied on survey samples since then.

7. Rose M. Kreider and Renee Ellis, "Number, Timing, and Duration of Marriages and Divorces: 2009," Current Population Reports P70-125 (May 2011): 19, table 10, www.census.gov/prod/2011pubs/p70-125.pdf. This report is drawn from the Census Bureau's 2009 Survey of Income and Program Partic.i.p.ation, popularly referred to in the industry as SIPP. Although there are other surveys that have begun to be fielded in recent years, the SIPP remains a standard that numerous experts refer to regularly. Many experts have high hopes for the American Community Survey, but because it is a brand-new type of annual survey, it will be years before it will show us any meaningful, long-term trends.

8. Data for 2001 shows 70 percent of both spouses are in first marriages in Rose M. Kreider, "Number, Timing, and Duration of Marriages and Divorces: 2001," Current Population Reports P70-97 (February 2005): 12, table 8, www.census.gov/prod/2005pubs/p70-97.pdf. Data for 1996 shows more than 70 percent of both spouses are in first marriages in Rose M. Kreider and Jason M. Fields, "Number, Timing, and Duration of Marriages and Divorces: 1996," Current Population Reports P70-80 (February 2002): 19, www.census.gov/prod/2002pubs/p70-80.pdf.

9. Arthur J. Norton and Louisa F. Miller, "Remarriage Among Women in the United States: 1985," Current Population Reports P-23, no. 169 (December 1990), www.census.gov/hhes/socdemo/marriage/data/cps/P23.169.1990.Report1.pdf. In this report, they delineated divorce and widowhood rates; the actual divorce rate for first marriages in 1985 was 23.1 percent. (This is calculated from page 2, table A: divorced after first marriage/ever married 17,142/73,971 = 23.17.) That means approximately 76.9 percent were either still married or widowed. Subtracting out those widowed will probably bring the number down to roughly at or below the 72 percent rate of today. Note that the 1975 and 1985 surveys are a different type of survey from the 1996, 2001, and 2009 SIPP surveys.

10. Links to the 1996, 2001, and 2009 SIPP (and the technical citation information) can be found at www.census.gov/hhes/socdemo/marriage/data/sipp/index.html.

11. Glenn Gritzon, "Super Composite for 20122013 of the Family Needs Survey Findings," FamilyLife, Little Rock, AR, 2013, www.familylife.com/FNS.

12. Although this is simply a rough estimate-since it applies the rate of widowhood from a completely different data set to the Census Bureau data-the end number of 20 percent may be a bit conservative. The eight-point drop due to widowhood was found in surveys that were primarily among churchgoers, who a.n.a.lytically are less likely to have various health-risk factors (such as smoking) that lead to mortality at younger ages. Thus, among the general population, the rate of widowhood could potentially be even greater than the eight percentage points found in the church population; so the actual first-marriage divorce rate could be lower than 20 percent.

13. There are multiple other well-researched studies that have found different ratios that, for various reasons, are not good nationally representative averages for the total population. For example, a 2009 study, "The Marriage Index," from the Inst.i.tute for American Values-led by an excellent team of researchers-shows in the table "Marriage Index: Monitoring the Nation's Leading Marriage Indicators" that only 61.2 percent of first marriages are intact, which on the surface could imply a quite high 38.8 divorce rate. But in addition to not knowing which marriages ended in death rather than divorce, this particular survey only sampled those between the ages of twenty and fifty-nine. In other words, it included the worst divorce years of those aged fifty to fifty-nine years (the baby boomer generation has the highest divorce rate) and didn't include the better data from the older women. (See www.americanvalues.org/search/item.php?id=44.) 14. As with many of these surveys, getting the correct "ever divorced" number took a bit of calculating. On the Census Bureau's "Marital History for People 15 Years Old and Over by Age and s.e.x: 2009" (Kreider and Ellis, "Number, Timing, and Duration: 2009," 16, table 6), the percent of women "ever divorced" is just 22.4 percent. But that is of the total female population surveyed (123,272 women) and not of women ever married (which was 89,742). Calculating 22.4 percent of the total population delivers the number 27,613, which, when divided by women ever married, brings the real ratio of those ever divorced to 30.8 percent.

15. Kreider and Ellis, "Number, Timing, and Duration: 2009," 16, table 6 (derived percentage as noted above).

16. The final 2012 "ever divorced" number was provided directly in personal correspondence with Tom Smith, November 20, 2013. For general survey information, Tom W. Smith et al. "General Social Surveys, 19722012," National Opinion Research Center, University of Chicago, March 2013, http://www3.norc.org/GSS+Website.

17. Norval D. Glenn, "With This Ring: A National Survey on Marriage in America," National Fatherhood Initiative, 2005, 32, http://blog.fatherhood.org/with-this-ring-survey. (The survey was conducted 20032004.) 18. Barna Group, "New Marriage and Divorce Statistics Released," March 31, 2008, www.barna.org/barna-update/family-kids/42-new-marriage-and-divorce-statistics-released#.UnfR8-Ao6Uk.

19. Knights of Columbus / Marist Poll Survey, 2010. Poll data first appeared in Carl Anderson, Beyond a House Divided: The Moral Consensus Ignored by Was.h.i.+ngton, Wall Street, and the Media (New York: Doubleday, 2010), 101. Poll data was provided to authors by Knights of Columbus via e-mail, April 18, 2013.

20. Kreider and Ellis, "Number, Timing, and Duration: 2009," 16, table 6, the birth cohort among women ages sixty to sixty-nine. Note that among the fifty to fifty-nine age cohort, the percent ever divorced was calculated as 41 percent. But that too gives a misleading picture. See the FAQ section for more on the "gray divorce" issue among baby boomers.

21. Beyond the "ever divorced" method discussed here, the Census Bureau has also attempted a completely different way of looking at and trying to estimate the divorce rate, as shown in figure 5 ("c.u.mulative Percentage of Ever-Married Women Divorced from First Marriage by Race and Ethnicity and Duration of First Marriage: 2009") in the SIPP report (Kreider and Ellis, "Number, Timing, and Duration: 2009," 15). This method delivers a c.u.mulative estimate that roughly 40 percent of first-time marriages have ended in divorce by forty years of marriage. However, in our judgment, the complex c.u.mulative methodology used on the data provided by survey takers delivers a result that contradicts the actual "ever divorced" numbers for those same people, shown in SIPP table 6. Table 6 shows a significantly lower divorce rate average for those survey takers who could have (by age) reached a fortieth anniversary. (Combining the age groups of those sixty and above, only 29 percent had ever been divorced, and even adding in a few younger boomers who got married young wouldn't change that percentage much.) In a November 14, 2013, conversation, the Census Bureau official who did the full SIPP report explained that figure 5 was an alternative way to try to get a handle on these complex divorce numbers. However, because the calculations for this method were not published, we were unable to examine them. (As far as we can tell, this c.u.mulative method has not been widely used by the demographers and sociologists who work in this field.) The official also explained that figure 5 was not primarily used to estimate the divorce rate but to show the difference in divorce between racial groups. Bottom line, while this c.u.mulative method might be helpful to show the differences in divorce between racial groups, it does not appear to deliver a number consistent with the actual "ever divorced" numbers published in table 6.

22. Among most experts, this fact is uncontested. However, a recent working paper by Dr. Sheela Kennedy and Dr. Steven Ruggles of the University of Minnesota theorized that because of widespread reporting problems (see next footnote), the divorce rate may not actually be continuing to drop. But whether or not the earlier trend has plateaued, the divorce rate is certainly far down from its peak. We have added a discussion on this in the FAQ section of the book.

See Sheela Kennedy and Steven Ruggles, "Breaking Up Is Hard to Count: The Rise of Divorce and Cohabitation Instability in the United States, 19802010" (working paper 2013-01, Minnesota Population Center, University of Minnesota, Minneapolis, April 2013), www.pop.umn.edu/sites/www.pop.umn.edu/files/WorkingPaper_Breaking_Up_April2013.pdf.

23. Even this number of reported divorces is fraught with problems, since there is such a wide swing in the consistency of reporting of the number of divorces at the local, regional, and state level from year to year. For example, a given city hall might report divorces one year and not the next, and that happens all over the country. But in aggregate, it is still the only number that has been tracked for years that gives us anything like a trend from the same sources, year after year, without changing how the number is calculated.

24. A related way of looking at this trend is the "refined divorce rate," which is the number of divorces per one thousand married women (rather than per one thousand people). The refined divorce rate should be more meaningful because it compares divorces to those married, but unfortunately we haven't found any source that has used the same data set for "number of married women" from the 1960s through today. So comparing the numbers would mean you were looking at one type of number in this decade and another type of number in a different decade, thereby meaning it's not a true trend. Several researchers have switched to using the American Community Survey (ACS), which added a question about divorce in 2008. This will eventually give us great data but at the moment is not helpful for a trend, especially when trying to compare today's divorce rate to the divorce rate peak in the 1979 to 1981 period. However, we can estimate the refined divorce rate using a chart created by sociologist Bradley Wright (obtained by personal e-mail); he calculates that the refined divorce rate in 2009 was 16.4. In other words, roughly sixteen married women out of every thousand got divorced that year. That refined divorce rate number peaked at 22.8 in 1979 and has been steadily falling ever since, declining more than 28 percent overall; it is therefore in the same ballpark as the crude divorce rate drop of 32 percent.

25. US Census Bureau, "Births, Deaths, Marriages, and Divorces," Statistical Abstract of the United States: 2012, 65, table 78, www.census.gov/prod/2011pubs/12statab/vitstat.pdf. For crude divorce rates from 20002011: Centers for Disease Control and Prevention, National Center for Health Statistics, "National Marriage and Divorce Rate Trends: Provisional Number of Marriages and Marriage Rate: United States 20002011," www.cdc.gov/nchs/nvss/marriage_divorce_tables.htm.

26. Manning, Brown, and Payne, "Two Decades," 22.

27. Manning, Brown, and Payne, "Two Decades," 6.

28. Scott M. Stanley, Galena K. Rhoades, and Howard J. Markman, "Sliding vs. Deciding: Inertia and the Premarital Cohabitation Effect," Family Relations 55 (2006): 499509, https://app.box.com/s/59fd9e71728ef08658be. This study found that couples who were living together without being officially engaged were more likely to divorce if they did get married. In general, various negative results in the relations.h.i.+p when couples live together before marriage (especially before getting engaged) are well doc.u.mented and known as the cohabitation effect. Dr. Stanley and others have found, for example, that these couples experience a higher risk of divorce later, as well as depression and domestic aggression. Couples who cohabit also tend to have less commitment (especially among men), satisfaction, and quality in the marriage.

29. Joshua R. Goldstein, "The Leveling of Divorce in the United States," Demography 36, no. 3 (August 1999): 410. "Divorce rates peak during the fourth year for both first marriages and remarriages," http://ccutrona.public.iastate.edu/psych592a/articles/Goldstein_1999.pdf.

30. Andrew J. Cherlin, "In the Season of Marriage, a Question. Why Bother?," New York Times Sunday Review, April 27, 2013, www.nytimes.com/2013/04/28/opinion/sunday/why-do-people-still-bother-to-marry.html?_r=0.

31. Rose McDermott, James H. Fowler, and Nicholas A. Christakis, "Breaking Up Is Hard to Do, Unless Everyone Else Is Doing It Too: Social Network Effects on Divorce in a Longitudinal Sample" (working paper, October 18, 2009), 38, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1490708.

32. Scott Stanley, e-mail message to authors, October 23, 2013. The data he references on baby boomers is from Susan L. Brown and I-Fen Lin, "The Gray Divorce Revolution: Rising Divorce Among Middle-Aged and Older Adults, 19902010," Journals of Gerontology Series B: Psychological Sciences and Social Sciences 67, no. 6 (2012): 73141, http://ncfmr.bgsu.edu/pdf/Susan%20L.%20Brown/file108701.pdf.

33. Interestingly, Dr. Stanley's study also reviewed several other cohabitation studies and found that people who are "more traditionally religious are less likely to cohabit prior to marriage."

34. Andrew Cherlin, "Demographic Trends in the United States: A Review of Research in the 2000s," Journal of Marriage and Family 72, no. 3 (June 2010): 40319, www.ncbi.nlm.nih.gov/pmc/articles/PMC3293163. "Nearly all studies suggest that the lifetime probability of [marriage] disruption is between 40% and 50%." He includes Drs. Raley, b.u.mpa.s.s, Schoen, Stevenson, and others as references.

35. Teresa Castro Martin and Larry L. b.u.mpa.s.s, "Recent Trends in Marital Disruption," Demography 26, no. 1 (February 1989): 49, www.jstor.org/stable/2061492. They estimate a 56 percent first-marriage divorce rate from the census of June 1985.

36. From the 1980s projections of a 50 percent average divorce rate, the average projection today is 40 to 50 percent-in other words, 45 percent on average. That means researchers' projections have only come down about 10 percent, but the actual crude divorce rate has dropped 32 percent. (Note: up until 2009, the last year we have good comparison data, the refined divorce rate dropped 28 percent.) If the projections had followed anywhere close to the same trend, demographers would be predicting more like a 34 percent divorce rate for a newly married couple.

37. Paul Amato, phone interview with the authors, February 19, 2013.

38. Dana Rotz, "Why Have Divorce Rates Fallen? The Role of Women's Age at Marriage" (working paper, Mathematica Policy Research Inc., Harvard University, 2011), http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1960017.

39. Dana Rotz, e-mail message to authors, May 29, 2013.

40. Scott Stanley, personal meeting with author, Denver, CO, January 10, 2013.

41. The study surveyed 1,304 married people-652 married couples-over the course of 2011 and 2012. To keep the surveys independent, anonymous, and as candid as possible, each person took the survey separately from their spouse. The spouses did not see each other's surveys. There were two different types of data sets. The first type were independent surveys, conducted during 2011 and 2012 either on paper or with direct-response keypads with 796 people (398 married couples) in group meeting places with a high concentration of married couples, including a church wors.h.i.+p service, marriage conferences, a cruise for married couples, and weekend retreats. Among these independent surveys, largely of churchgoers (and probably including a disproportionate number of struggling marriages), 34 percent were very happy, 37 percent happy, and 29 percent struggling. The second type of data set was a nationally representative survey conducted for us by the research firm Decision a.n.a.lyst with 508 married people (254 married couples) between March 23, 2012, and April 2, 2012. Although the survey was nationally representative for demographics such as age, race, and geographical area, it was not representative for marital happiness. We had to inflate the number of struggling couples to get a statistically significant (large enough) sample. So we tried to come relatively close to the percentages we'd found in the independent surveys. Within the Decision a.n.a.lyst survey sample, 39 percent were very happy, 38 percent happy, and 22 percent struggling. See the memo in the research section of www.shaunti.com by survey designer Dr. Chuck Cowan for The Surprising Secrets of Highly Happy Marriages for a further explanation of our methodology.

42. These are the results of our independent surveys and do not include the Decision a.n.a.lyst numbers. As noted above, that survey was not nationally representative for marital happiness.

43. Smith et al., "General Social Surveys."

44. The last year currently available for viewing on the GSS web-site (as of this writing) is 2008. The happiness percentage for married people in 2008 resembled the 2012 numbers: "very happy" was 62 percent, "pretty happy" was 35 percent, and "not too happy" was 3 percent (numbers rounded up).

45. Glenn, "With This Ring," 36, http://blog.fatherhood.org/with-this-ring-survey. From appendix B, "Technical Description of Survey": "The survey for this report was designed to be representative of the United States resident population age 18 and older and was conducted by telephone by the Office of Survey Research at the University of TexasAustin in December of 2003 and January and February of 2004.... The response rate was 89 percent according to the most commonly used method of calculating response rates for telephone surveys (number of interviews/number of interviews + refusals), and there were 1,503 completed interviews. The questionnaire was designed by Dr. Norval Glenn in consultation with advisors at the National Fatherhood Initiative."

46. Despite our many requests for more detail, this study was one of the very few on which we were forced to rely on abstracts and articles and were not able to get closer access to the study or researchers. However, because GfK is a credible survey company and this survey was showing a different and lower number than the other studies, we felt it was important to include it as a counterpoint. Here is the citation information and methodology: Michele Kimball, "Poll: Most Marriages Are Happy," Divorce360.com, www.divorce360.com/divorce-articles/statistics/us/poll-most-marriages-are-happy.aspx?artid=268. "The poll was conducted by the independent research firm for Divorce360.com. GfK Roper polled more than 1,500 people in September. The polling sample was made up of about 55 percent women and 45 percent men. The margin of error for the study is plus or minus 2.6 percent." The abstract did not say the year, but from the summary, it looks as though it was 2008.

47. Note that these numbers total only 95 percent. Because (as noted above) this was one of few studies for which, despite many requests, we were unable to get direct access to the study and researchers, we are forced to rely on published information from the sponsor of the study. That published information does not provide full details, including what categories make up the missing 5 percent.

48. Anderson, Beyond a House Divided, 1067. Matthew St. John at Knights of Columbus provided the following additional information on how the Knights of Columbus / Marist National Poll July 2010 survey was conducted: "This survey of 2,029 adults was conducted July 9 through July 13, 2010. Adults eighteen years of age and older residing in the continental United States were interviewed by telephone. Telephone numbers were selected based upon a list of telephone exchanges from throughout the nation. The exchanges were selected to ensure that each region was represented in proportion to its population. To increase coverage, this landline sample was supplemented by respondents reached through random dialing of cell phone numbers. The two samples were then combined and balanced to reflect the US Census results for age, gender, income, race, and region. Results are statistically significant within plus or minus 2.2 percentage points. There are 485 Catholics. (Knights of Columbus is a Catholic organization.) Results for this subset are statistically significant within plus or minus 4.5 percentage points. The error margin increases for cross-tabulations."

49. Survey of Marital Generosity, 20102011. Special data run from Brad Wilc.o.x for Shaunti Feldhahn to cross tabulate by marital happiness categories, August 28, 2013. The survey was conducted with 2,230 married men and women between the ages of eighteen and fifty-five who had children at home. The logistic regression model also adjusted for partic.i.p.ants' age, education, household income, and race/ethnicity. Although couples were grouped in similar categories to my Surprising Secrets of Highly Happy Marriages study, since this was a different survey conducted in a different way, I have t.i.tled the resulting categories "very happy," "happy," and "less than happy" to distinguish the results from my own categories of "highly happy," "mostly happy," and "so-so and struggling."

50. Christine Johnson et al., "Marriage in Oklahoma: 2001 Baseline Statewide Survey on Marriage and Divorce," Oklahoma Marriage Initiative. Findings for this report are based on telephone interviews conducted with a statewide sample of 2,323 adults. The sample consisted of 2,020 adults from randomly selected households and 303 randomly selected current Medicaid clients. Based on the total sample, one can say with 95 percent confidence that the error attributable to sampling and other random effects is plus or minus 2.02 percentage points.

51. The GfK Roper poll is an example of a survey that asked a much more specific question about what percent of the time the marriage was happy. This is probably why the number (75 percent) was a bit lower and perhaps more realistic.

52. In addition to the Marist and National Fatherhood Initiative surveys noted here, other polls have found similar numbers. For example, an official CBS News poll, "Love and Marriage," has found that since they began conducting the survey in 1995, 90 to 93 percent of married Americans would marry their spouse all over again. (CBS News Poll, "Love and Marriage: January 2931, 2010," www.cbsnews.com/htdocs/pdf/Poll_Jan10dLove.pdf.) 53. Marist Poll, "Do You Think You Married the Right Person, or Not?," http://maristpoll.marist.edu/86-most-americans-hitched-to-the-right-wagon-but. Nature of the sample: This survey of 1,004 US residents was conducted June 17 through June 24 of 2010. Residents eighteen years of age and older were interviewed by telephone. Telephone numbers were selected based upon a list of telephone exchanges from throughout the nation. The exchanges were selected to ensure that each region was represented in proportion to its population. To increase coverage, this landline sample was supplemented by respondents reached through random dialing of cell phone numbers. The two samples were then combined. Results are statistically significant at plus or minus 3.0 percent. There are 530 residents who are married. The results for this subset are statistically significant at plus or minus 4.5 percent.

54. Glenn, "With This Ring," 34.

55. Linda J. Waite et al., Does Divorce Make People Happy? Findings from a Study of Unhappy Marriages, Inst.i.tute for American Values, 2002, 4, 11, https://docs.google.com/viewer?url=http://americanvalues.org/catalog/pdfs/does_divorce_make_people_happy.pdf.

56. There has been some lively debate among sociologists and psychologists about whether the Waite study in some way focuses more on couples experiencing easier-to-solve marital discord rather than deeper marital distress. (See S. R. H. Beach and F. D. Fincham, "Spontaneous Remission of Marital Discord: A Simmering Debate with Profound Implications for Family Psychology," Family Psychologist 19 [2003]: 1113.) However, in our basic review of Waite's study and supporting material that she e-mailed to us directly, it is clear that her methodology was rigorous and that she is including couples of all categories, including couples whom we would consider to be in marital distress. Furthermore, her a.n.a.lysis was based on how the couples themselves rated their marriages over time; she did not a.s.sign them rankings. So although we are aware of the controversy (in part spurred by the question of whether the benefits of sticking with it are "too good to be true"), in our judgment the Waite study appears to be well done.

57. Waite et al., Does Divorce Make People Happy?, 5.

58. Waite et al., Does Divorce Make People Happy?, 5.

59. The following are results from currently married persons who at some point thought their marriage was in trouble and considered divorce, who were asked "Are you glad you are still together?" Among those married seven years or less, 79 percent were glad, 15 percent were unsure or had mixed feelings, and only 6 percent were not glad. Among those married eight years or more, 95 to 97 percent were glad (depending on number of years married), 1 to 5 percent were unsure/mixed, and 0 to 2 percent were not glad. (See endnote 50 in this book and page 28, table 22, of the report.) 60. Unpublished song by Danny Oertli, November 17, 2010. Used with the permission of Danny Oertli.

61. The article was about the CDC's newly released National Survey of Family Growth data from 2006 to 2010, declaring that 48 percent of women were divorced by their twentieth anniversary-without mentioning that those numbers were entirely of women who married very young. For an adjustment to those numbers, see the FAQ section.

62. This chapter will focus solely on churchgoers, rather than those attending wors.h.i.+p services of other faiths. Although several studies (including my own) surveyed those attending Jewish or Muslim services, for example, the sample size was in many cases too small to be statistically significant. However, from what we can tell from the smaller numbers, these conclusions appear to apply to those who attend wors.h.i.+p services in other religious settings as well.

63. Some studies, for example, look at wors.h.i.+p attendance for any faith; others just look at Christians. Some define "regular attendance" as weekly; others as twice a month or more. Other studies simply look at "religious affiliation"-maybe with attendance, maybe without. Still others bypa.s.s wors.h.i.+p attendance and study other actions such as praying together, studying the Bible, and discussing G.o.d's will in marriage. And there are several studies, such as Barna's, that look at religious beliefs, not practices. Some of the confusion in researching this topic has also revolved around the publication dates. Several high-profile academic studies to come out in the last five to seven years are still a.n.a.lyzing NSFH data from 1987 to 1994.

64. Special Report for Shaunti Feldhahn, 2008 OmniPoll, Barna Group, Ventura, CA, 2013.

65. Since the Barna Group wasn't studying church attendance per se, they didn't ask questions such as "How often do you attend services?" Instead, they asked something like "Have you done one of these things in the last seven days?" with one option being whether the person had been to church. Although that will include a few extra people (those who were in a church out of curiosity or at a wedding, for example), it will also exclude a few regular attenders (for example, those who were on vacation and missed church that week). So it still is probably a fairly good proxy for those who actually attend church regularly.

66. Barna cla.s.sifies those who hold specific beliefs as "evangelical" (respondents were not asked to describe themselves as such). Specifically, the Barna Group says evangelicals are defined as "people who said they have made a personal commitment to Jesus Christ that is still important in their life today and who also indicated they believe that when they die they will go to Heaven because they had confessed their sins and had accepted Jesus Christ as their savior." This group also includes those who "say their faith is very important in their life today; believing they have a personal responsibility to share their religious beliefs about Christ with non-Christians; believing that Satan exists; believing that eternal salvation is possible only through grace, not works; believing that Jesus Christ lived a sinless life on earth; a.s.serting that the Bible is accurate in all that it teaches; and describing G.o.d as the all-knowing, all-powerful, perfect deity who created the universe and still rules it today. Being cla.s.sified as an evangelical is not dependent upon church attendance or the denominational affiliation of the church attended." (See www.barna.org/barna-update/family-kids/42-new-marriage-and-divorce-statistics-released#.UsDeWPRDtng.) 67. Another area of good news has to do with the rate of marriage itself. While sociologists continue to be concerned by decreasing marriage rates (in part due to people living together), that trend is also better in the average church. According to our tabulations of Barna's 2013 data, while the overall rate of marriage did decline, 79 percent of weekly church attenders are or have been married compared to just 66 percent of nonattenders.

68. W. Bradford Wilc.o.x and Elizabeth Williamson, "The Cultural Contradictions of Mainline Family Ideology and Practice," in American Religions and the Family: How Faith Traditions Cope with Modernization and Democracy, ed. Don S. Browning and David A. Clairmont (New York: Columbia University Press, 2006).

69. Unlike most other surveys, this survey asked questions designed to get at nuances about attendance and affiliation. The couples were asked their religious preference and how often they attended a religious service, among other things. If the respondents answered Protestant, they were then asked, "What specific denomination is that?" This kind of detail tends to illuminate even clearer patterns about drops in the divorce rate by attendance than those found in the Barna survey (which, as noted, was not designed to focus on nuances about attendance).

70. James A. Sweet and Larry L. b.u.mpa.s.s, "The National Survey of Families and Households-Waves 1 and 2: Data Description and Doc.u.mentation," Center for Demography and Ecology, University of WisconsinMadison, 1996, www.ssc.wisc.edu/nsfh/home.htm. The first two waves of more than thirteen thousand people were completed between 1987 and 1994, with ten thousand follow-up interviews in the second wave. Many different researchers have a.n.a.lyzed significant samples from the initial wave and/or follow-up interviews (from "Sample Design," www.ssc.wisc.edu/nsfh/design.htm).

71. The NSFH/Wilc.o.x a.n.a.lysis clearly found a much larger drop in divorce among those who attend church (compared to those who don't) than the Barna Group did. The real answer today is probably somewhere between the two. The NSFH is comprised of older data, but the survey was also much larger and is considered more precise than the Barna survey, particularly on this topic. Barna surveys, as noted earlier, did not ask partic.i.p.ants whether they regularly attended church or did other religious activities but simply asked if they had been in church the last seven days, thirty days, and so on. (Our tabulations for this book included those who had been in church the last seven days.) When we compare the Barna and NSFH data to other studies we have, it appears that the drop today is likely somewhere in between the Barna average drop of 27 percent and the NSFH/Wilc.o.x average drop of 50 percent.

72. W. Bradford Wilc.o.x, "Is Religion an Answer? Marriage, Father-hood, and the Male Problematic," Research Brief No. 11 (New York: Inst.i.tute for American Values, 2008), www.americanvalues.org/search/item.php?id=20.

73. Although the "Cultural Contradictions" study itself lists the reductions by various divorce groups and denominations, it does not list one final overall average for all Christians. In correspondence with us and in his paper "Is Religion an Answer?," Dr. Wilc.o.x repeatedly referred to an average drop of 50 percent before controlling for other factors, and an average drop of 35 percent after controlling for those factors.

74. One interesting finding was that before controlling for socioeconomic factors, mainline Protestants had lower rates of divorce than various evangelical groups that, traditionally, have had a greater stigma against divorce, such as evangelical Protestants and black Protestants. In "Is Religion an Answer?" Dr. Wilc.o.x explained that "evangelical Protestants and black Protestants face somewhat higher divorce rates because they are more likely to hail from working-cla.s.s and poor communities where economic struggles often stress marriages."

75. Margaret L. Vaaler, Christopher G. Ellison, and Daniel A. Powers, "Religious Influences on the Risk of Marital Dissolution," Journal of Marriage and Family 71 (November 2009).

76. Christopher G. Ellison, professor of sociology, University of Texa.s.san Antonio, e-mail message to authors, April 28, 2013.

77. Gritzon, "Super Composite." The Family Needs Survey is highly specialized, asking church members 180 questions to understand their degree of faith commitment, how their family relations.h.i.+ps work, what their habits are, and so on. See www.familylife.com/FNS.

78. In part because of the demographic makeup of some of the churches, and in part because it was a long survey that young couples with kids couldn't take the time to complete, 48 percent of survey takers were over age fifty and 68 percent were over age forty.

79. David McLaughlin, "The Role of the Man in the Family," David McLaughlin audio series, www.disciples.h.i.+plibrary.com/search.php?a=1&e=0&m=0&n=0&p=0&s=series&t=SERIES&st=series&ss=Role%20of%20the%20Man%20in%20the%20Family.

80. Special a.n.a.lysis from "When Baby Makes Three: How Parenthood Makes Life Meaningful and How Marriage Makes Parenthood Bearable," The State of Our Unions 2011 (Charlottesville, VA: National Marriage Project at the University of Virginia, 2011), 3132, figure 13, www.stateofourunions.org/2011/when-baby-makes-three.php. Special a.n.a.lysis of couples in each happiness category, for Shaunti Feldhahn.

81. Here's a bit more detail: Where 97 percent of married people had happy marriages and 68 percent said they were very happy, that 68 percent average wasn't nearly as informative as its parts. It turned out that among those who never attend religious services, only 52 percent were very happy, whereas among those who attended regularly (at least monthly), the number was 72 percent or higher.

82. Peter J. Larson and David H. Olson, "Spiritual Beliefs and Marriage: A National Survey Based on ENRICH," Family Psychologist 20, no. 2 (2004): 48, www.apa.org/divisions/div43/news/NewsArchives/Spring04TFPfinal.pdf.

83. Robert Lewis and Tim and Lea Lundy, Marriage Oneness, FamilyLife (2010), www.marriageonenessprofile.com.

84. Marriage Oneness Survey of 7,700 married individuals (3,850 couples) tabulated by PREPARE/ENRICH, October 10, 2013. Surveys conducted as part of the "Marriage Oneness Profiles" for partic.i.p.ants in the Marriage Oneness video study by FamilyLife (http://lifeready.com/marriageoneness/). Levels of closeness are based on a validated and robust couple typology scoring algorithm developed by PREPARE/ENRICH. It requires both individuals to rate their relations.h.i.+p healthy across a range of the Marriage Oneness topics in order to be categorized as "highly connected."

85. Leo Averbach, "The High Failure Rate of Second and Third Marriages: Why Are Second and Third Marriages More Likely to Fail?," Psychology Today, for Mark Banschick, The Intelligent Divorce, February 6, 2012, www.psychologytoday.com/blog/the-intelligent-divorce/201202/the-high-failure-rate-second-and-third-marriages.