Difference in Divorce

[soflabbwlvr]

Hmm, good point. It would of course be useless in court if the judge wont see it. But maybe there are other legal avenues where such evidence could be used?

I'm thinking about those "Alienation of Affection" suits that are so often use in revenge stories, but which I have never actually encountered in real life.

Or a lawsuit against the company in whose office your wife banged her co-worker for failing to uphold their own moral guide lines - another frequently used plot device that I've never seen for real.

The reason for that is so-called "heart balm" statutes have been abolished in most states. In Florida, since around 1968.

The morality clauses do not grant a private right of action to third parties. They are designed to protect the company from sexual discrimination lawsuits, and to protect the company from adverse publicity. Non-parties to the employment relation have no actionable interest in enforcing the policy.

 

[carlieplum]

In the State of Alaska the state will step in and pay the mother the missing child support and then go after the dad, who now all of a sudden is in dept to the state rather than his ex. And they're relentless. Regardless how bitchy your ex wife is, they'll make her seem like a pussycat in comparison.

(And if you ask me how I know this, I'll plead the fifth... )





True that. Custody is probably the most frequent point of conflict. Unfortunately this has evolved into a situation where making child abuse accusations against the father is part of the standard arsenal of battle techniques. It's efficient too since the authorities always err on the side of caution and there aren't any consequences for the accuser if the allegations proves to be false.

Child support is a gender-neutral issue, calculated on who had the kid more and who makes more money. The state will go ofter a deadbeat mom and garnish her paycheck too.

 

[Tx Tall Tales]

Interesting thing I ran across:
A 2004 Stanford Business School study compared outcomes in states that adopted no-fault divorce versus those that did not. It found:

20% reduction in female suicide after 20 years, none for men
33% reduction in domestic violence against women (after a rise in other states vs. a drop in no-fault states)
Reduction in the domestic murder rate for women, none for men

Study authors argued that in part, men were encouraged to behave better because they knew it would be easy for their spouses to divorce and find another partner

I found this extremely interesting, enough to read the actual report, and the studies it was based on. I think the most interesting part was left out of the summary.

In suicides, the mean for women was 54 suicides per million vs. 202 for men. From 1964 to 1996, women suicides went down 10% while men stayed the same. 49 per million to 202. (Don't know where the 20% came from, the average effect varied from 5.4% to 9.7% according to their charts, after application of the controls.)

For Domestic Violence, again the men started on the short end of the stick:

Overall Violence(a)
Husband to Wife 11.7%
Wife to Husband 11.9%

Severe Violence(a)
Husband to Wife 3.4%
Wife to Husband 4.6%

The Husband to Wife violence declined by 1.7% (about a 16% drop) in states with no-fault divorce from 1976 to 1985, the only 2 years that were sampled by a survey where families talked about how they resolved family disputes. These are not hospital records or police reports, but self professed family issues.

Even more confusing is the comparison. 30 states vs NY, TN, MS, that had not instituted No-fault divorces. (The other states and had not instituted no-fault divorces had no data for 1976 and were excluded)

For Murder, it gets more bizarre. The period was from 1968 to 1994. The rate for wives killed by spouses, dropped 10% in the 8 years before no-fault was introduced and 2% in the following 20 years. For men the 'none' reduction was an INCREASE of men murdered by their wives of 12.3%

In all cases, the men were killed more often, abused more often, and committed suicide more often than their spouses. After no-fault, the difference got even worse.

Looks like no-fault divorce created a lot of angry women.

I just love how people manipulate statistics. What do they say about the 3 kinds of lies? "Lies, Damned Lies, and Statistics."

 

[StrangeLife]

The reason cod that is so-called "heart balm" statutes have been abolished in most states. In Florida, since around 1968.

"Heart balm statutes" - I like the name

But isn't a marriage in essence a contractual obligation?

Thus it would seem logical that - regardless of how the divorce and custody issues are resolved - you should be able to successfully launch a civil lawsuit against your ex for breach of contract demanding reimbursement for all expenses caused by her actions.

Child support is a gender-neutral issue, calculated on who had the kid more and who makes more money. The state will go ofter a deadbeat mom and garnish her paycheck too.

That's how it should be and how it is in theory. In reality though your chromosomes invariably weigh in on the decisions and being equipped with the right genitalia will off-set the balance significantly in your favor.

In other words, all things being equal the mom is most likely to get custody for no other discernible reason than she's the mom. At least that has been my observation.

 

[Bramblethorn]

While most of the U.S. has adopted no-fault divorce, that doesn't stop the parties from attempting to insert the issue into the proceedings. The two biggest causes of divorce are money problems and infidelity. In cases of infidelity, the innocent spouse may spend hundreds or thousands of dollars on private investigators prior to consulting an attorney. These P.I.s are under no obligation to advise their clients that whatever they discover will be largely inadmissible and wholly irrelevant. In most cases, infidelity only becomes a genuine issue when it is accompanied by and can be traced to a dissipation of marital assets. Parties often try to raise infidelity in custody cases, but 99% of the time it isn't allowed.

Annoyingly enough, while "vanilla" infidelity doesn't seem to be much of an issue for custody, consensual nonmonogamy has repeatedly been held against people in custody cases. Some since overturned on appeal, some not. eg:

http://beyondstraightandgaymarriage.../09/pennsylvania-court-rules-that-father.html

 

[Bramblethorn]

Sorry, this is gonna be long. Source, for the lazy: https://gsbapps.stanford.edu/researchpapers/library/RP1828.pdf

In suicides, the mean for women was 54 suicides per million vs. 202 for men. From 1964 to 1996, women suicides went down 10% while men stayed the same. 49 per million to 202. (Don't know where the 20% came from, the average effect varied from 5.4% to 9.7% according to their charts, after application of the controls.)

Looking at Table 2 (p. 28), what they're saying is that twenty years after introduction of no-fault the suicide rate for women was down by about 11.2-18.7% ("almost 20%" depending on which column you like best), but because the effect ramped up gradually, averaged over those first 20 years the reduction is about half that. (See Figure 1 on page 32 for a graphical version of some of that data: it's plain to see that the final point at the end of 20 years is lower than the average across those 20 years).

The usual story with suicide is that women make more attempts, but men use more lethal methods so have more 'successes'.

Editing the next bit out of order:

In all cases, the men were killed more often, abused more often, and committed suicide more often than their spouses. After no-fault, the difference got even worse.

Men were killed more often (p. 24), but not by their spouses. "Women killed by spouses" overall was 7.3 per million per year, about 33% higher than the figure of 5.5 for men killed by spouses. The overall rate of homicide for men is about 3x that for women, but that's mostly due to homicide by non-family-members.

For Murder, it gets more bizarre. The period was from 1968 to 1994. The rate for wives killed by spouses, dropped 10% in the 8 years before no-fault was introduced and 2% in the following 20 years. For men the 'none' reduction was an INCREASE of men murdered by their wives of 12.3%

Hmm. I'm not seeing that. Are you looking at table 5 there? (p. 31) That's the only place I can see which resembles those numbers, but that's not what it's saying. By my reading:

In that table, the first column shows the change in homicide rates (assume I'm talking about by-spouse rates for the rest of this comment) for no-fault states relative to the change in non-no-fault states. So for instance, if the homicide rate goes up by 5% for states that introduced no-fault in that period, and 10% for states that didn't, a simple analysis would suggest that no-fault makes a difference of -5%.

The figure of "12.3%" for men killed by spouses doesn't mean that homicide rates for men went up by 12.3% in no-fault states; it means that it went up by 12.3% relative to what happened in other states.

However, that's only a simple analysis. Homicide rates can be influenced by a bunch of other factors - e.g. if a no-fault state got hit by a major economic downturn, you'd probably expect to see an increase in spousal homicide rates for that state. If you ignore that factor, you'd end up misinterpreting the rise in homicide as a consequence of no-fault.

To deal with that problem, they've attempted to control the analysis for confounding factors (rather a lot of them, listed in a footnote). Column 2 shows what the effect of no-fault looks like when you allow for those controls. When you do that, the "woman killed by spouse" effect looks a little larger (increases to 12.6%) and "man killed by spouse" drops quite a bit, to 3.9%.

(I'm not going to get into columns 3-4 of that table; I think it's about correcting for limitations of the control technique by evaluating what the controls do to non-intimate homicide, but I'd have to read more closely and nobody's paying me for this ;-)

It's also important to note the margins of error on those numbers. The figures in (parentheses) show estimated standard error for each number; in general you can expect the true figure to be within one s.e. of the estimate around 66% of the time, and within two s.e.s around 95% of the time. So their listed effect of "3.9% (9.0)" for men killed by spouses means "we're about 95% sure the effect is somewhere between -14.1% and +21.9%".

In other words: there's so much 'noise' in the data for men killed by spouses that it's impossible to tell what no-fault did. It might have caused a big reduction, might've caused a big increase, might've done nothing at all.

 

[Tx Tall Tales]

Sorry, this is gonna be long. Source, for the lazy: https://gsbapps.stanford.edu/researchpapers/library/RP1828.pdf

Looking at Table 2 (p. 28), what they're saying is that twenty years after introduction of no-fault the suicide rate for women was down by about 11.2-18.7% ("almost 20%" depending on which column you like best), but because the effect ramped up gradually, averaged over those first 20 years the reduction is about half that. (See Figure 1 on page 32 for a graphical version of some of that data: it's plain to see that the final point at the end of 20 years is lower than the average across those 20 years).

Look at table in figure 3. The maximum decline in any one age group is less than 8%. The authors describe how they massaged the numbers, and say "... our preferred estimate, suggesting an average decline in female suicide of around eight percent, and a long run decline that is perhaps double that."

Not 20%. Cherry picking. Find enough variables to distort the data, and pick the one that proves what you want. This is what I was trying to point out. That and the fact that men are the ones committing suicide, 400% more often, and no-fault did not change that.

No mention of the studies that show divorced women have suicide rates three times that of married women.

The figure of "12.3%" for men killed by spouses doesn't mean that homicide rates for men went up by 12.3% in no-fault states; it means that it went up by 12.3% relative to what happened in other states.

I believe that's was the point of the whole analysis, the difference in states with no-fault, vs. those without.

To deal with that problem, they've attempted to control the analysis for confounding factors (rather a lot of them, listed in a footnote). Column 2 shows what the effect of no-fault looks like when you allow for those controls. When you do that, the "woman killed by spouse" effect looks a little larger (increases to 12.6%) and "man killed by spouse" drops quite a bit, to 3.9%.

(I'm not going to get into columns 3-4 of that table; I think it's about correcting for limitations of the control technique by evaluating what the controls do to non-intimate homicide, but I'd have to read more closely and nobody's paying me for this ;-)

It's also important to note the margins of error on those numbers. The figures in (parentheses) show estimated standard error for each number; in general you can expect the true figure to be within one s.e. of the estimate around 66% of the time, and within two s.e.s around 95% of the time. So their listed effect of "3.9% (9.0)" for men killed by spouses means "we're about 95% sure the effect is somewhere between -14.1% and +21.9%".

In other words: there's so much 'noise' in the data for men killed by spouses that it's impossible to tell what no-fault did. It might have caused a big reduction, might've caused a big increase, might've done nothing at all.

I understand what you're saying, and agree. Numbers are manipulated, trying to guess what the reasons and effects are, and you are right, there are many. In the end, it is interesting that they choose to chart 'intimate' Homicide, vs. Spousal Homicide, when we're talking about Divorce. They show a lot of numbers, then say:

"Figure 4 confirms the initial findings of a decrease in women murdered in the period
following the passage of divorce law reforms. However, the timing evidence is somewhat
worrying, and the reader is left to judge whether the decline in homicide pre-dated the law change
to an extent that undermines our results. This raises the possibility that our regression results may
be picking up the effects of some alternative phenomenon that pre-dated divorce law reform."

In essence, there was a large decline in female homicides BEFORE the no-fault happened, and in actuality, the decline SLOWED, after the introduction of no-faults divorces.

Again, my point is, someone selected the data they liked, and now it's being published all over the place. Blame Wikipedia on that.

Numbers reported, without an explanation, are what makes statistics so damning. Seriously, their analysis of Domestic violence, self-reported, with only two years available, the first a decade into the introduction of no-fault, is highly suspect. Add in the effect that the control group consists of three states, NY, TE, and MS, and it gets uglier. And what they do report, supports my final statement. Men got abused more, and continued to get abused more.

According to the Bureau of Justice Statistics, (http://www.silentwitness.net/sub/violences.htm) From 1994 to 2010, the overall rate of intimate partner violence in the United States declined by 64%, from 9.8 victimizations per 1,000 persons age 12 or older to 3.6 per 1,000. Is this the effect of no-fault divorce 50 years later, or is the decrease in domestic violence due to other factors, awareness perhaps.

Here's an interesting point, also from the Bureau of Justice Statistics: "Females living in households comprised of one female adult with children experienced intimate partner violence at a rate more than 10 times higher than households with married adults with children and 6 times higher than households with one female only."

If no-fault increases the number of divorces, (2.3% in 1964, 5.3% in 1981, 3.6% in 2012) it would seem that it is responsible for a huge spike in intimate partner violence.

I guess I'm a cynic. I think most of the people doing these things have an axe to grind. I'm suspicious of the numbers, and in this case, I believe rightfully so. I'm not looking to argue what the data is, without full access to the researchers information, it would be useless. I'm sure you agree that statistics can be made to show whatever you want them to.

Again, domestic violence has been in decline for over 60 years. Using this study to credit it to No-Fault divorces seems ridiculous.

 

[Bramblethorn]

(re. claim of almost 20% drop)

Look at table in figure 3. The maximum decline in any one age group is less than 8%.

Yep. In fact, although the chart design is confusing, I think all but the very last chart (the one marked "RHS") are supposed to be using the scales marked on the left-hand margin, in which case the maximum decline shown for an individual age group would be about 3% for the 35-44 group.

But it's important to understand what those graphs mean. They're not saying "rates within 35s-44s declined by 3%". They're saying that the decline in 35-44s accounted for a 3% decline in the TOTAL rate for women of all ages. The table is explained on page 12:

"Figure 3 shows our results by age group for women, mapping the estimated response of suicide rates following the change in divorce regimes, where the analysis in Figure 1 is repeated separately for women in each of eleven different age groups. These age groups comprise unequal shares of the population, and so in each case coefficients are scaled by their share of the US population, allowing these figures to be added to yield the aggregate effect (shown in the bottom right panel)."

As for the discrepancy between the 12% indicated by Figure 3 and the ones given elsewhere... I'm a bit puzzled there. Figure 1 looks consistent with column f2 of table 2 (final effect -16% with a standard error around 3%) and that makes some sense, because they've said that f2 is their preferred analysis. So far so good.

The note at the bottom of Figure 3 says: "Bottom right panel: Top line reproduces aggregate result from Figure 1." The shape looks very similar, but the numbers are different - this one ends up at around -10%. My best guess is that something's gone wrong with the scale on that panel.

Not 20%. Cherry picking. Find enough variables to distort the data, and pick the one that proves what you want.

If they'd just run four different analyses and arbitrarily chosen the one that gave the most impressive effect, I'd agree with you: cherry-picking. But from the explanation, that doesn't seem to be what they did.

"We test the sensitivity of our specification to a range of controls, including [long list of stuff]. While we find that some of these controls are significant explanators of the suicide rate, column 2 shows that they barely change the estimated effect of unilateral divorce. This column represents our preferred estimate, suggesting an average decline in female suicide of around eight percent, and a long run decline that is perhaps double that. Weighted least squares results are also broadly similar (column 3). We add state-specific time trends in column 4, finding that their inclusion causes the standard errors to increase. For women, the specification including state-specific time trends is not precisely estimated enough to reject either a null that the pattern of coefficients follows that shown in columns 1-3, or a null of no effect."

In other words: the first three analysis methods gave very similar results (estimates from 16.4% to 19.3% decline, all with SEs around 3%) and the fourth had so much error that it wasn't saying anything at all. Without having seen the details, I'd guess it led to an overspecified problem; if you try to consider too many explanatory variables, and you don't have enough data points, you end up with massive errors.

So they discarded that one, and of the remaining three they picked the one with the smallest estimated effect. 16.4% with a SE of 3.3% - if you're trying to summarise that for the introduction, "around 20%" seems reasonable, if a little vaguer than I would've chosen. Anybody who wants to know just how precise "around 20%" might be can read the rest of the paper.

This is what I was trying to point out. That and the fact that men are the ones committing suicide, 400% more often, and no-fault did not change that.

No mention of the studies that show divorced women have suicide rates three times that of married women.

It's already a 34-page paper; a researcher who tries to include everything that might possible be related will never finish writing. In this case the point is "what effect does no-fault divorce have?" so I wouldn't expect them to give a lot of space to things that don't directly bear on that.

I wouldn't be at all surprised to hear that divorcees have higher suicide rates, but correlation does not imply causation. (By way of parallel: accident victims who go to the ER are more likely to die than those who don't, NOT because the ER is a dangerous place to be, but because those who are worst injured are most likely to go there.) Which is why this paper goes to some trouble to check whether causation seems plausible - e.g. by testing whether declines might have started before no-fault came in. (Answer: doesn't look like it.)

I believe that's was the point of the whole analysis, the difference in states with no-fault, vs. those without.

Not quite, no. The point is to look at what difference is caused by no-fault - which is estimated by comparing no-fault and non-no-fault states, and then compensating for other factors known to affect rates.

I understand what you're saying, and agree. Numbers are manipulated, trying to guess what the reasons and effects are, and you are right, there are many. In the end, it is interesting that they choose to chart 'intimate' Homicide, vs. Spousal Homicide, when we're talking about Divorce.

They do look specifically at spousal homicide ("By Spouse" rows of table 5) but they also examine broader categories ("by family", "by known"). The reasons for that are discussed on pages 18-19:

- Spouse-spouse relationships aren't the only ones that might be affected by changes in the law, and changes the law might affect people's marital status, which could confound the results.

For instance: if we find that NFD is associated with a big decrease in "murdered by spouse", one interpretation is that the availability of NFD makes the dynamic safer. But another explanation is that people are still getting murdered at the same rate, just that they're getting divorced first so that the person who murders them is no longer a "spouse". Or perhaps they're reacting to NFD by killing their kids instead?

In order to catch that sort of thing, it's reasonable to keep an eye on "killed by family" and "killed by known", as well as "killed by spouse". They do acknowledge: "The defect of the broader measures is that the treatment group is defined to include many relationships that are not affected by the treatment of unilateral divorce. The defect of narrower measures is that police classifications of victim-perpetrator relationships as “spousal” are likely to have changed over time, in a way that is correlated with family law regimes..."

They show a lot of numbers, then say:

"Figure 4 confirms the initial findings of a decrease in women murdered in the period
following the passage of divorce law reforms. However, the timing evidence is somewhat
worrying, and the reader is left to judge whether the decline in homicide pre-dated the law change
to an extent that undermines our results. This raises the possibility that our regression results may
be picking up the effects of some alternative phenomenon that pre-dated divorce law reform."

In essence, there was a large decline in female homicides BEFORE the no-fault happened, and in actuality, the decline SLOWED, after the introduction of no-faults divorces.

Looking at Figure 4, the data is noisy enough that I wouldn't be willing to say for sure whether the decline preceding introduction of no-fault is large compared to the decline afterwards. But it certainly does look as if it started before introduction of NFD; large or small, that does undermine the suggestion of a simple causal relationship there.

My best guess (completely unsubstantiated) is that the same social/political climate that made it possible for NFD to go through in the first place was also influencing relationships through other channels.

Numbers reported, without an explanation, are what makes statistics so damning.

Yep. Problem is, nobody wants to quote 30-odd pages of caveats and definitions and clarifications; they don't even want to read it. They mostly skim it for something that seems sensible (i.e. coincides with their biases) and quote that nugget. I don't think most people intend to misrepresent sources, but unconscious bias is hugely powerful.

Seriously, their analysis of Domestic violence, self-reported, with only two years available, the first a decade into the introduction of no-fault, is highly suspect.

Why is "self-reported" a problem?

They do acknowledge that having only two comparison points available limits the power of the analysis, but by my understanding that should be reflected in the standard errors published with the data.

Add in the effect that the control group consists of three states, NY, TE, and MS, and it gets uglier.

I count eleven, not three. Quoting from page 15 and footnote on page 16:

"...we can compare changes in violence rates among our thirty-seven states that constituted the “no-fault revolution” with two alternative control groups: the five states that are yet to adopt unilateral divorce (AR, DE, MS, NY and TN), and the nine states whose pre-existing regime involved unilateral divorce (AK, LA, MD, NC, OK, UT, VA, VT and WV)... The 1976 survey did not sample from all states, and hence we are forced to omit the following states from our analysis: AK, AR, DC, DE, HI, IA, KY, MA, ND, NH, NM, NV, RI, SD, WY. "

That only leaves MS, NY, and TN from the first control group, but it also leaves LA, MD, NC, OK, UT, VA, VT, and WV from the second group, for a total of eleven. Because the point of the analysis is to look at how DV rates change when NFD is introduced, a state that already had NFD pre-1976 is just as valid a control as a state that didn't have it by 1985.

Again, the size of the control group would be reflected in the standard errors reported.

And what they do report, supports my final statement. Men got abused more, and continued to get abused more.

Fair point, and one which merits attention. It just isn't the focus of this particular study.

(Note also that while male-on-female DV appears to be about as common as female-on-male, male-on-female is more likely to turn deadly, going by the spousal murder stats. That might reflect different modes of violence, or just differences in physical power.)

Here's an interesting point, also from the Bureau of Justice Statistics: "Females living in households comprised of one female adult with children experienced intimate partner violence at a rate more than 10 times higher than households with married adults with children and 6 times higher than households with one female only."

If no-fault increases the number of divorces, (2.3% in 1964, 5.3% in 1981, 3.6% in 2012) it would seem that it is responsible for a huge spike in intimate partner violence.

Not so fast. You have correlation there between single parenting and intimate partner violence there, but one of the first rules of statistics is to be wary about assuming correlation = causation. There are all sorts of ways that data could be interpreted.

The fact that women who are married to their partner experience less violence than those who aren't could mean that something about being married discourages violence (in which case, easier divorce ought to increase DV rates). But it could also mean that women encounter a mix of abusive and non-abusive partners, and are more likely to marry a non-abusive one - in which case marriage is a consequence of violence (or rather, lack of it) and not a cause. This sort of complication gets back to why the authors of this study looked at other categories besides spousal homicide, in order to identify whether a supposed "improvement" is just a matter of shifting violence into a different category.

The fact that women living alone with kids experience more violence than those living alone without kids could mean that looking after kids puts more stress on relationships, or it could mean that it's harder to leave a bad relationship when there are kids involved.

I guess I'm a cynic. I think most of the people doing these things have an axe to grind. I'm suspicious of the numbers, and in this case, I believe rightfully so. I'm not looking to argue what the data is, without full access to the researchers information, it would be useless. I'm sure you agree that statistics can be made to show whatever you want them to.

If the methodology is bad, sure. That's why researchers are expected to publish their methodology, so you don't have to take it on trust. I'm not sure but I think all the data they used here is publicly available, so if you have a good stats background it'd be possible to reproduce it (at least broadly) and see whether you get the same results.

Again, domestic violence has been in decline for over 60 years.

Yep, which is why the study goes to a lot of trouble to differentiate potential effects of NFD from overall trends; it's not crediting NFD for all of that decline.

 

[soflabbwlvr]

Kudos. Here we have a discussion with differing viewpoints, and there is an absence of vitriol and name calling. No allegations of hidden agendas. I'm inspired.

 

[carlieplum]

I'm also impressed by the erudition! I'm getting quite the education.