There is a common theme in the media suggesting that baby boomers are divorcing in greater numbers now that they are in or approaching retirement. The most recent example I’ve seen is from Casey Dowd at Fox Business (H/T Empathologicalism): Why So Many Baby Boomers are Getting Divorced. The breathless headline is accompanied by the obligatory image of an older woman fed up with her husband. The actual words need not be written:
So long boring loyal dude!
This is all standard fare for those selling divorce. Rule number one is you only sell divorce to women. Women eat this up, which is why every form of women’s entertainment has divorce fantasy as a staple. Selling divorce to men would be in bad taste, as this would leave fatherless children and more importantly men wouldn’t stand to be pandered to so shamefully. There are a number of other rules, including giving the impression that everyone is doing it, especially older women. If you can convince the public that grey divorce is rampant you can have maximum impact destabilizing and demoralizing all marriages. Why stick through difficult times if late life divorce is inevitable? You may as well quit now.
The problem is the line between news and entertainment is frequently blurred. Because of this we have a rash of headlines suggesting there is an explosion in divorce around retirement. This is quite simply not the case. Here is what Mr. Dowd writes in the Fox Business article:
The divorce rate among boomers has jumped recently and that number is only expected to climb.
He cites statistics from the National Center for Family & Marriage Research (NCFMR) at Bowling Green State University in making this claim. With a bit of searching I found these excellent charts* from the NCFMR. After looking at the data referenced as well as the working paper from the same group on the topic it is clear that the statement referenced above is incorrect in both of its assertions. The divorce rate among boomers has fallen, not increased, and it is expected to fall further, not climb. In addition, this and other articles leave the false impression that late life divorce is relatively common.
As I’ve shown previously and the charts linked above confirm, divorce is much less common as couples age. The changes under discussion are large percentage changes in very small numbers. For example, look at the divorce rates per 1,000 married couples age 55-64 in figure one in the chart linked above. In the US in 1990, there were roughly** 5 divorces per 1,000 married couples in this age bracket. Looking at the same age bracket today, it has doubled to 10 per 1,000 married couples. This value of 10 is very small compared to the roughly 30 divorces per 1,000 couples for those now in their twenties and early thirties. While there will be a few more grannies batting their bifocaled eyelashes on the dating scene, this is more of a trickle than a flood. You can get a sense of perspective from a chart I put together using similar but slightly different data than what the NCFMR shows in the charts linked above. Here is how divorce risk changed by age when looking at women in the US in 2009:
More to the point, if you are a boomer your divorce risk has dropped dramatically over the last 20 years. You might be wondering how this could be given the doubling stats I referenced above, but the answer is simple. Twenty years ago boomers were twenty years younger and experiencing much higher divorce rates than they are now. Here is what happened to divorce rates for boomers and silents using the NCFMR Figure 1 data**:
Here is the data compiled in chart form:
While there is something interesting happening, there is no jump in divorce rates for any of the cohorts of boomers and silents measured in the data referenced. All cohorts experienced dramatic reductions in divorce rates during this time, and are expected to experience more of the same. What has happened is a bending of the curve of divorce distribution across time. I noted the same basic phenomenon in the UK when considering the divorce implications for young marriage here (see final chart). The simplest answer for the US data seems to be that boomers and silents simply divorce at much higher rates than the generations before or after them. As they move through the age brackets, this would account for making the curve steeper in 1990 and less steep in 2010. Part of this could be that these cohorts of boomers and silents divorced and remarried more than previous cohorts did when they were younger, and second and third marriages have much higher divorce rates. At any rate, this is something very different than the claimed jump in divorce rates as boomers and silents retire.
But a bending curve won’t sell copy, and it won’t create the kind of “everyone is doing it” sensation which sells divorce. So instead Mr. Dowd interviews an “expert” from a firm which sells divorce services, whispering:
Do you think these trends are going to continue? Is the 30-year itch the new seven-year itch?
*This data was compiled similarly to the data I compiled myself here. My data is for 2009 and theirs is for 2010 but the two still are extremely close.
**Their charts don’t have data value labels so I may be slightly off in reading their values. However, even if I’m off by one or two either way it doesn’t change the analysis.
See Also: Divorcée Retirement
If I’m reading the table and graph approximately right — say I’m in the Boomer cohort (b. 1956-1965) and married at age 25 in 1980. Then my cohort’s average annual rate of divorce was 33/1000 per year in the decade ending in 1990, X in the decade ending in 2000, and 19/1000 per year in the decade ending in 2010.
I’ll interpolate and estimate X as 26.
That would give an estimate of the percentage of my cohort that becomes divorced over those 30 years as 33% plus 26% plus 19%, or 78%.
Hmmm. That sounds way too high, even for Boomers.
I know there’s a lot of complexity that these back-of-envelope calculations miss. E.g. once a couple is divorced, they leave the “married” group for all subsequent years. And as mentioned, second marriages muddy the waters.
Or perhaps I’m looking at this all wrong?
mgwk,
The problem with those back of envelope calculations is the data here isn’t looking at one marriage cohort, it is looking at divorce rate point values over time for a birth cohort. People in each age bracket will have been married for different numbers of years. Part of this is some couples waiting longer to marry, and others are remarriages. If you want to see what it looks like by marriage cohorts check out the third and fourth charts here.
Dalrock, Thanks for those links — clears up my bafflement!
Thank ye sir, i must make me humble blog a more mathy place, why, cause i like it
Women replace traditional wisdom for consumerism & social supremacy, every time
Men replace nuclear families with child welfare & alimony & no fault divorce, governmental sanctioned fatherhood
Older married women, build social networks around their marriage
Younger women, dont have the social skills or customs or the network to stay vested in a marriage
Outside of an extended family, the society of a marriage acts as a surrogate for a real extended family
Without this extension of the marriage, a woman has no place in a marriage past her fertility
Younger women spend so much time socialising,spent on facebook, ludicrous lengthy phone calls in preparation for infertility
Infertility & its preparation plays a massive part in a womans youth
Infertility & its preparation plays a massive part in a womans youth
Which is why divorce & infertility ie abortion, is so effective on women
As it forces women to be in preparation for infertility in perpetuity
Infantalising maturity as consumerism, is always an easy sell for women who see infertility as failure, as opposed to maturity & adage
Perhaps a bit OT, but psychologist, researcher and writer Judith S. Wallerstein has died at age 90. Among her books was “The Unexpected Legacy Of Divorce” from 2000 or so.
Great post, the increase in each curve proves that there will be a tipping point down the line when divorce rates will reach a maximum and only those interested in genuine marriages will marry, lowering the overall marriage rate but reducing divorce rates.
It’s as if though marriage will become some kind of fitness test.
It’s confusing because the charts clearly show that divorce rates are both increasing and decreasing at the same time.
Divorce rates are increasing across generations – that is to say more recent generations are divorcing at a higher rate than previous generations.
However, divorce rates within a particular generation decrease over time.
The takeaway is that people marrying today are more likely than previous generations to end up divorced, for all of the various reasons discussed here at Dalrock’s and other blogs. Conversely, the longer a couple of any generation stays married, the greater the chance that they will continue to stay married, until death do they part.
Here is a interesting note- the numbers are very inaccurate. In review of data collecting it is severely hindered. The last comprehensive data collection was 1989-1990.
The following was taken from the http://www.cdc.gov/nchs/mardiv.htm#reporting website
“The collection of detailed data was suspended beginning in January 1996. Limitations in the information collected by the States as well as budgetary considerations necessitated this action. The most recent comprehensive analyses of detailed marriage and divorce data are published in Advance Report of Final Marriage Statistics, 1989-90 and Advance Report of Final Divorce Statistics, 1989-90.”
@Michael Singer
They only used data from Vital Statistics to calculate divorce rates for 1990. For 2010 they used the 2010 American Community Survey.
@BlackCat
Not exactly. Boomers and the cohort of silents referenced divorced at higher rates than the cohorts before them. But gen x and y are divorcing at lower rates than the boomers and silents did when they were that age. So divorce rates for younger age brackets went down between 1990 and 2010, while rates for older age brackets went up. Think of the boomers and silents as a wave of divorce moving through time.
Correct.
FYI, as mentioned the method, reliability of data,data collection, and accuracy has changed quite a bit. The 1990 numbers were defacto – the 2009 numbers are “extrapolated” from only 2 million people ( very small considering the overall population). Hence the numbers are grossly inaccurate compared to 1990.
Here is from the Marital Events of Americans: 2009 – http://www.census.gov/prod/2011pubs/acs-13.pdf on page 1 & 3.
“The American Community Survey (ACS) was started in the late 1990s to replace
the collection of data from the long-form questionnaire in the decennial censuses.
The U.S. Census Bureau mails a quartermillion ACS questionnaires every month to a nationwide sample. Follow-ups are conducted with nonrespondents to the mail questionnaire by phone and inperson interviews, collecting data from a sample of about 2 million households annually. The sample is then weighted to be representative of the nation’s population as a whole.
These new marital events items fill a void in the marriage and divorce data collected in the United States………..
Historically, data on marriages and divorces in the United States were collected from marriage and divorce certificates filed and collected at the state-level through the vital statistics system by the National Center for Health Statistics (NCHS). In 1996, the Centers for Disease Control and Prevention (CDC) and the NCHS decided to discontinue the collection of detailed state-level vital records data from marriage and divorce certificates. In the absence of up-to-date vital records information on marriages and divorces, the quality of U.S. marital events estimates has diminished. Consequently, the U.S. Department of Health and Human Services (DHHS) approached the Census Bureau about adding marital history items to the ACS in order to have reliable and valid data for researchers and policymakers alike as part of its Healthy Marriage Initiative.
With the resumed collection of marital events data on the ACS in 2008, reliable estimates of marriage and divorce rates, among other statistics, are available for the nation and states annually.
This report uses data from the 2009 ACS to describe the population in the United States
who recently experienced a marital event (marriage, divorce, widowhood). Included within this report are geographic, demographic, and historical profiles that describe the current state of marriage, divorce, and widowhood in the United States.”
“. . . every form of women’s entertainment has divorce fantasy as a staple.” This is the best quote in the article. The media sells divorce to women like Satan sold forbidden fruit to Eve. Only our modern day “Eves” dont need to look at us with a longing stare for us to get screwed over, they just need to fetch the papers that society tells them will make their dreams come true.
@ Michael Singer
Pointing out the limitations of the data is valid, but what is your conclusion? Are you saying that because the data is extrapolated it is worthless and we should go on the media’s portrayal of their feelings on the subject? Simply casting aspersions on the data, and therefore Dalrock’s analysis, is an underhanded way of taking a position without taking a position. If you disagree with Dalrock, it would be more beneficial to the discussion if you actually came out and said so and outlined the reasons why. That is the type of debate that allows clarity and definite conclusions to be reached.
I hold the position that data trumps no data for identifying trends, unless that data is proven to be completely fabricated. From your posting, the DHHS feels that the data they provide is statistically significant and reliable as in: “With the resumed collection of marital events data on the ACS in 2008, reliable estimates of marriage and divorce rates, among other statistics, are available for the nation and states annually.” I’m going to go with their analysis until you can convince me otherwise.
“Why _So Many_ Baby Boomers are Getting Divorced”
As a numbers guy, I am strongly bothered by this trend of conjuring up “trends” out of thin air with no stats to back it up. The Hymowitz posse has done the same thing, with their “too many men are dropping out and playing video games” without any justification about how to measure or determine “too many.”
“Too many” is a wholly subjective judgment, based (in Hymowitz’s own words) on the experiences of her 20-something interns and underlings who date losers, and the appearance of schlubby guys in successful movies.
What was hilarious about Kay’s WSJ article was that it tried to visually lay out this “problem” using a photo collage of a bunch of dudes and a woman in a business suit with an atrocious haircut. It was a photographic Rorschach test, no doubt women thought this girl was a super catch, meanwhile us guys went “whaat?”
Back on the actual point: Truth be told it’s a very feminine way of thinking, where whatever you “feel” or “perceive” trumps the objective reality.
@ Dalrock:
“But gen x and y are divorcing at lower rates than the boomers and silents did when they were that age. So divorce rates for younger age brackets went down between 1990 and 2010, while rates for older age brackets went up. Think of the boomers and silents as a wave of divorce moving through time.”
Do you think gen x and y are divorcing at lower rates because these generations are marrying at (albeit slightly) lower rates? I think we know the numbers on marriage are creeping downward and age at first marriage is creeping upward. But we also know that we hear about the never marrieds and the “where are all the good men” crowd because they are the most vocal.
“The simplest answer for the US data seems to be that boomers and silents simply divorce at much higher rates than the generations before or after them.”
This seems to be the money quote. It also seems to bear out the effects of the social, political and cultural upheavals that generation fomented and lived through — the sexual revolution; cheap, safe and effective birth control, no-fault divorce; Roe v. Wade; and individual rights. It also seems to coincide with the visible rise of political power of special interest groups (women, gay rights, reproductive rights, and so forth).
@Singer
I’m with Draggin, I’m not clear where you are running with this.
@ draggin, deti, and dalrock.
The numbers are FAR GREATER than what is being captured in the data and the problem is FAR WORSE than the report suggests.
Dalrocks previous post of “Whistling through the graveyard?” had a link to show that divorce was declining from Susan Walsh on “Your Chances of Divorce May Be Much Lower Than You Think”. In it she shared some stats on divorce rates from the 2010 State of Our Unions report by The National Marriage Project at the University of Virginia which used extrapolated data from 45 states data vs. actual data from 50 states. It is sloppy sloppy sloppy statistics of which can be used- knowing the severe limitations of the accuracy.
I strongly suspect the defacto numbers can be found at the state level and complied easily but no one has bothered, has “the balls” to collect it and present the real findings.
@Michael Singer
I haven’t studied the ACS methodology, but a sample size of two million is huge. They still could have managed to screw up the sampling, but the size of the sample doesn’t look like a problem to me. There will be some margin of error as there always is with a sampling, but what other study have you seen with this kind of sample size?
One interesting thing is their measurements show that divorce rates per 1,000 married have been flat since 1990 at 19, while the Marriage Project stats show a decline. You pointed out on the other thread that the Marriage Project data lacked several big states, including California. This data is a representative sample of the entire country, so I’m inclined to think it is more accurate. A good sample is better than a biased census. I’m going to email the Marriage Project folks with the question of why their numbers differ so far from NCFMR (using ACS survey data). Interestingly they look like similar organizations. Both The Marriage Project and NCFMR are university sponsored projects looking at marriage trends.
@Deti
It would make sense that when the trend is toward later marriage those who still marry early are more committed to marriage than the population at large. I think there is another component that the groups with the highest divorce rates simply don’t marry as often any more. It might also be that after seeing the trail of wreckage the boomers and silents left in their wakes gen xers and yers are taking marriage more seriously (on the margins).
@Dalrock. In regards to 2 million in terms of 360 million is actually pretty small data set for EVERYTHING parameter (marriage, divorce, death etc….) They use 1 form for everything -ONE.
For comparison – consider the populations of Los Angeles 3.8 million, Chicago 2.7 million, Houston 2.1 million and New York 8.2 million alone.
To use a sample size of 2 million is to take a large city like Houston or Chicago and have represent the entire nation.
Consider how it was done and what is recorded on the US Government website – they openly admit their own shortcomings !!!!!!!!!!!
“…the quality of U.S. marital events estimates has diminished”
“Historically, data on marriages and divorces in the United States were collected from marriage and divorce certificates filed and collected at the state-level through the vital statistics system by the National Center for Health Statistics (NCHS). In 1996, the Centers for Disease Control and Prevention (CDC) and the NCHS decided to discontinue the collection of detailed state-level vital records data from marriage and divorce certificates. In the absence of up-to-date vital records information on marriages and divorces, the quality of U.S. marital events estimates has diminished”.
http://www.census.gov/prod/2011pubs/acs-13.pdf
They have moved from actual to projected and when they do use actual – 5 states are left out.
By changing the method of collection on two fronts and hence understating the results- what is presented is just the “tip of the iceberg”.
Btw, I am not disagrement with the post. What I am suggesting is – the data sets are severely flawed (major red flags as with the data in the previous post) and the problem is far worse given that “no fault” historically increases the divorce rate.
@Michael Singer
Not if they did their sampling correctly.
I think you are misreading that quote. They are explaining why they added the questions to the ACS. The full data isn’t being gathered as it used to be. They are helping plug the gap by including marriage and divorce questions on their survey. It makes sense.
Thanks. I am trying to be careful as well. I’m not wedded to any one source of data. I’m trying to make sense of and share the best data I can find. If there are flaws in the data that is something we need to be aware of. With this post I was pointing out that the data the article referenced didn’t back up the assertion being made. In fact, it proved the opposite. With that said, I’m very curious how the Marriage Project folks will respond regarding their divorce stats and what the NCFMR calculated with the ACS data (I just emailed them). NCFMR is saying divorce rates per 1,000 married couples has remained flat for 20 years at 19, while the Marriage Project is saying it has declined to 16.4. This is a large difference in what appears to be the same exact metric. If you know of a better source of this statistic or the others we are discussing, I’d love to know about that/those as well. Lets put all of the available data on the table along with the explanations.
Just in case you are interested, the NCFMR has some other data on divorce rates by education which they share here. This is different from Marriage Project results I shared in the last post in some surprising ways. However, this isn’t apples to apples since the Marriage Project data we just looked at was looking at a specific marriage cohort (late 1990s) while the NCFMR data is looking at first marriage divorces which happened in 2010. Essentially they are doing the same breakdown by education that this post looked at by age, but they are only looking at 2010. I’ll hold off on doing a post with this NCFMR data for a bit to give the Marriage Project folks some time to make their case for or against the NCFMR/ACS data. In the meantime, if you are interested you can see a list of other documents the NCFMR have put out here.
@ Dalrock. Reporting sources have move from actual (very accurate) of the entire population to telephone/mail of a small sample size.
With that being said, the accuracy of what is reported and the actual number that are useable (ie washout) of the 2 million is a even a smaller population with the highest degree of inaccuracy.
What would be worthwhile and valuable beyond belief would be the gathering of actual state data (all 50 states). With accurate data in hand, the only only problem left would be presenting the data.
Sorry, I cannot agree with Michael Singer’s description of sampling techniques, based only on less than a 100% sample. That is simply not how it works, and I thus assume he has not studied statistics at all. A sample of 2 million out of 300 million produces numbers very close to correct 99% of the time. Data if all data would be valuable more ways then one, including it would cost 150 times more, with little increase of accuracy.
I mean no disrespect as such, Mr. Singer, but you almost certainly do not understand sampling techniques.
California, Colorado, Indiana and Louisiana don’t report divorces. The population total from those states is ~54 MM people. Were these statistics you state normalized for these nonreporting populations, some of which have traditionally high rates of divorce?
Anon 70
I concur
Dalrock, this is OT, but Ed Hurst has written an excellent blog post on feminism and marriage. Perhaps you would care to showcase it in an upcoming post of your own:
http://jehurst.wordpress.com/2012/06/19/husband-and-lord
The NCFMR statistics say that 18 out of every 1,000 people get divorced — you will notice that most statistics about how high the rate is say that X out of every X MARRIAGES end in divorce. Thus, the NCFMR statistics do not include repeat offenders who participate in multiple failed marriages. I am almost certain that this number has gone up over time. Whether this works for or against your argument I will leave up to you.
@collapseofman
The 18 figure is per 1,000 marriages. From the pdf with the charts:
@mjay
This is the apparent problem with the Marriage Project calculation which places the current number of divorces per 1,000 marriages at 16.4. The NCFMR based their calculation on the American Community Survey (ACS) which is nationwide, and came up with a current value of 18 (I think I mistakenly wrote 19 somewhere above). I’ve emailed the Marriage Project asking them explain why they think their partial census of divorces is better than the ACS full survey. I’ll share their explanation if/once I receive it.
@Anonymous age 70- allow me to offer a reframe.
A sampling size of 2 million mailers / phone of which there are non responders, misreported data(ie lying), incomplete, there is a very high degree of inaccuracy and washout vs. actual state reported data (defacto). Just because one samples 2 million doesnt mean one will get 2 million responses (defacto).
Acquiring state data isn’t that hard- it is public record.
I openly admit I dont know much about statistics – I do know a little bit about studies, populations, washout, study design, and a sample size of 2 million out of 365 million with the methodology used vs. actual reported data is comparing a that is of the highest accuracy to that of the lowest accuracy and hence the least valid. To compare the two in a graph is a JOKE !
It is a different methodology and study design.
FYI, meta-analysis is the least reliable method – this is a well known fact.
Btw, it would be interesting to know what was the total number out of 2 million given the washout, non repsonders, incomplete, and sampling error etc…. Chances are it is a pretty small number given it wasn’t reported.
Allow me to suggest the problem is far WORSE than what is being presented. Th current methodology is joke – The data is EASILY available.
If the (putatively better) data is easily available, why don’t you provide it?
Michael Singer says: “I openly admit I dont know much about statistics – I do know a little bit about studies, populations, washout, study design, and a sample size of 2 million out of 365 million with the methodology used vs. actual reported data is comparing a that is of the highest accuracy to that of the lowest accuracy and hence the least valid.”
At least you are honest. But you really need to study statistics and sampling some more before you make such broad claims denigrating this sample size. A properly constructed sample of 2 million can approximate the real data very well at considerably less cost. That sample construction is key; if they’ve done it well, the data will be reasonably accurate. That is the question that must be asked before we have license to trash the survey.
“Allow me to suggest the problem is far WORSE than what is being presented.”
Where are you getting that from? How are you arriving at that conclusion? I’m certainly open minded to your argument. If you can make it stick.
“. . . every form of women’s entertainment has divorce fantasy as a staple.”
I suppose Tolstoy was not into women’s entertainment. Anna Karenina was about the ultimate in a divorce fantasy horror. I wonder how Oprah ever recommended it.
So long boring loyal dude!
Oh, I am sure that the divorcing older baby boomer lady thinks she is exciting, but in reality she is probably just loud. Maybe he was boring, but he was loyal. So what is the score?
Guy 1
Women 0
Funny how the world twists things around.
@Random Angeleno- I’ll touch on this one more time – My background is with Biotech reviewing clinical studies with world thought leaders. I have learned a couple things over the years about studies, populations, study design, robust and presenting them (I was being a bit self effacing when it come to statistics in studies).
To compare factual reported data (high degree of accuracy) to a self reported mail/phone questionnaire is a joke
To use actual data and self reported data in the same graph is bad statistics.
With every study there is a washout, non responder, error, and response rate etc. A study of 2 million just got very very small amidst 365 million. Why do you think they dont publish this interesting little tidbits of information ?
Btw, this is a super easy epidemiology study since the facts are EASILY available. To conduct a ineffective study of 2 million out 365 million is a more expensive and the least effective route.
The report has some holes in it – the main fault is study design and the data used – it is simply too inaccurate. This is blatantly obvious.
Shalom
Btw, vast difference between verified interpolated data and unverified extrapolated data.
Btw, this is a super easy epidemiology study since the facts are EASILY available.
Just FYI, from one guy’s perspective, the whole “the data is easily available and so much more accurate” schtick doesn’t pass the most basic BS test. If it is so EASILY available, put up or shut up. Otherwise it just comes off as an ignorant rant.
FWIW, at one point a bunch of folks seemed to think I knew a thing or two about probability and statistics in a professional capacity. Also FWIW, I find your original explicit self assessment on the subject of statistics quite plausible.
Pingback: Conventional wisdom on the trend in US divorce rates may be about to change. | Dalrock
Hey Dalrock,
Logged on to my computer today and saw this:
Clear heads needed as senior divorces soar, sourced from a US News and World Report story. Opens with: Sad to report, but divorces among older Americans are surging. That’s saying something in a country that is already the world leader in divorces. And with baby boomers–already no strangers to divorce–entering middle age at a fast clip, the divorce trend may well become even more pronounced.
Sounds like you need to educate some folks on the facts. And the story reads less like “how to avoid a divorce in later life” than a “As a financial advisor, here’s my advice on how to square away details on a divorce”. sales pitch.
Yea they are on again at fox too
http://video.foxbusiness.com/v/1722716439001/baby-boomer-divorce-rate-skyrockets/?intcmp=fbfeatures
the lawyer is outright blatantly selling divorce, you deserve to be haaaapy and freeeee
Great retort! This was a really interesting read – and I read a lot of the linked articles too, thank you for sucking my time up! 🙂 I love graphs and charts.
I think it might also be worth considering the element of fertility to the mix. I know my friends are very obsessed about their family life and it is everything they have ever worked towards. For them to start again would be a kick in the teeth, but for them to break up their family when they are no longer able to create a new family would be truly tragic.
So yes, we need to directly compare the same age cohorts of all generations against each other to get a full picture, especially GenX and GenY. I would also be interested in any data showing which gens pursued separation help beforehand and other professional advice to save their marriage. We might find a very different story that the blase approach offered in the original article.
Pingback: Does Shaunti Feldhahn’s rosy divorce data prove that no fault divorce is working out pretty well after all? | Dalrock