As readers are aware from my recent articles, I contend the campaign for same sex marriage has lead to a deceleration in the rate-of-increase in-the-non-marital birth ratio [ROI-NMBR][1] in the United States. As I see it, Dr. Stanley Kurtz and I agree on one thing: there is a causal link between SSM and the ROI-NMBR.
It has come to my attention that this causal link is disputed.
In this article, I will provide hyperlinks to arguments for and against the causal link. I also invite readers to participate in a poll.
I will begin by presenting the case supporting the causal link. First, Dr. Kurtz, who has earned a Ph. D. in social anthropology from Harvard, says there is a causal link between SSM and ROI-NMBR. If such an elite educated person says a causal link exists, must it not be so?
Others also believe it. As recently as May 30, three days days after I published my first article showing the causal link for the American data, and two days after it was picked up by MarriageDebate.com, Maggie Gallagher, a white, educated elite opponent of SSM appeared to endorsed the cause and effect relationship and provided a supporting theory, saying:
Shortly afterward, a blogger Michael Sellitto posting at MarriageDebate.com also agreed there could be a causal link, writing:
I believe that summarizes the arguments in favor of the causal link. Let us now turn to arguments against it.
Almost as soon as I posted my first article in this series, Julian said
Later on, Rachel Ann noticed I buried Julian’s observation deep in the footnotes of this article. Probably suspecting[a] my intention to evade Julian’s comments, she posted to endorse his point of view. This raised the visibility of the idea. Afterwards both Trish Wilson[b] and Don P. posted in support of Julian.
How can one fail to be suprised by all this support despite the fact that Julian has expended very little ink to support his claim. Certainly he has devoted much less ink than either Dr. Kurtz, Mr. Katz or I have![2]
These are not the only critics of the proposed causal link. Since my first article appeared, I have discovered people posting on other blogs also believe there is no causal link.
These include:
- Andrew Sullivan who ridiculed the idea there could be cause and effect relationship.
- Jason Kuznicki of Positive Liberty, who implied that Dr. Kurtz and I might be guilty of data mining. (Jason was kind enough to note that my American Data has a much more noticeable kink than Dr. Kurtz’s Dutch Data which has no apparent kink.)
-
Jari Koskiku “I live in Finland and I cannot see how gay registred partnerships have had any kind of effect on how straight people view marriage. None. There is not enough evidence to say to what extent gay civil unions strenghten the concept of marriage, but there is enough evidence that the reasons for changes in the institution of marriage stem from somewhere else.” and
- Mark Becker suggests:
As much as I appreciate Stanley Kurtz’s attempts to look at what’s happening in Europe now for some sign of what SSM means for the future, I feel that it’s way too soon to be able to make any definitive statements of cause and effect..
Most surprisingly of all, on June 4th, after my article appeared, Maggie Gallagher states:“Isolating the effects of legal changes using social science methods is difficult. If the mechanism is the cultural meanings of marriage, the consequences are likely to take a generation to uncover and they may be different in different countries.”[3]
Examining all the arguments and counter arguments, I have decided the we must resolve this issue through the democratic process. You may all vote in this poll which I have set up at yahoogroups. [4]
Naturally, you may all campaign for your personal theory in the comments boxes. [5]
(Note: I made some editorial revisions to reflect comments.)
==========================
End notes:
[1] As readers are aware, this is the parameter Dr. Kurtz has selected as best illustrating the effect of SSM on the health of marriage. However, the phrase is so cumbersome I wish to refer to this with the acronym ROI-NMBR.
[2] Based on his parsimonious use of ink, one might develop the impression that Julian, who seems to be somehow associated with an elite university, but so far has not indicated the acquisition of a Ph. D., thinks the idea correlation might not indicate causation has been widely aired and accepted. What are they teaching at the University of Chicago?
[3] Can it be that, in just a few days, Maggie Gallagher has changed her opinion, and now believes it may be difficult to identify the effects of legal changes using the available data? Possibly, she simply means that SSM might strengthen marriage in one country, as we see in the US, while destroying it in another country, which Dr. Kurtz suggests is happening in the Netherlands.
[4] To particpate in the poll, you must join the yahoo group. Unfortunately, Yahoogroups doesn’t let me open the polls to the world. I will explore other polling feature for future polls. Submit ideas for future polls in the comments section. (I have set up the groups so that I cannot read addresses, nor identify who voted for the different choices.)
[5] Suggestion for campaign logos welcome.==== Added on revision
[a] Rachel Ann has defended herself against the accusation that she has such a suspicious nature.
[b] Many of these people, have afterwards, suggested that they are not so much agreeing with Julian, as they are disagreeing with me. Trish suggested that this might be an example of a logical error... - Mark Becker suggests:
But it IS very difficult to deduce causality using empirical methods in social sciences. There are so many other variables changing at the same time, and it is quite likely that these variables stand in complicated relationships to each other. Consider these possibilities: Some variable x (say, promotion of single-sex marriages) is thought to have caused y (changes in out-of-wedlock births). The reason for this guess is that the two appear to be correlated over time.
But there are so many other reasons why two things appear to be correlated over time. One is pure chance. If both x and y keep changing up and down over time, then it is almost certain (if they are unrelated, i.e. not in a causal relationship) that at some points in time one is going up when the other is also going up, and so on. This would mean nothing.
Also, if x and y move together, it may be that x is not causing y, but that y is causing x. This is unlikely in my example case, but in many others it’s quite possible, or even that both x affects y and y affects x.
The explanation many favored is that there is some third variable, say, z (in our example, value change) which contributes to both x and y.
The two are not causally related but both have the same cause.
Other theories could be built. To distinguish between these using empirical data is very difficult. It requires both good methodology, and taking into account all the other possible variables that could have an effect. A few things can be done more easily. For example, a simple way to study the relationship between x and y is to test if using older data on one variable and younger data on another causes differences. If x indeed causes y, then last season’s y should not be related to this season’s x, but last season’s x should be related to this season’s y. Of course even here you’d have to control for any dependencies over time.
That’s why these things are debated. If the data gave an obvious answer, the debate would be over.
My personal view is that the two are not directly causally related, just because I can’t see what theoretical explanation would account for this.
“The explanation many favored is that there is some third variable, say, z (in our example, value change) which contributes to both x and y.
The two are not causally related but both have the same cause.”
To take and example from fluid mechanics, in a turbulent flow, the fluid velocity has an element of randomness. The velocity in one direction, w, (say “up”) and the velocity in another diretion, v (say “to the right”) can be correlated. In a shear flow, the two velocities are almost always correlated. However, no one thinks one “causes” the other. The reason has to do with the underlying turbulence structure
I believe, to relate to Echidne’s discussion her “x” would be my “v”, her “y” would be my “w” and her would call “z”. would be the underlying turbulence structure. (To here.
People spend a lot of time developing theories of the underlying turbulence structure and have, with some frequency, tried to determine which ones are supported by the data. So, data are collected and correlations are calculated and ….
I better stop now though. I wouldn’t want to give the impression this is campaigning for the other side
I think everyone should just vote. I’m going to go vote now..
Lucia:
Another problem with your posts is that, whether by design or carelessness, you repeatedly conflate the separate questions of whether it is possible that a causal relationship exists and whether such a relationship actually does exist. The fact that a relationship may exist does not mean that it does exist. Denying that we have enough information to determine whether the relationship exists is not not not the same thing as denying the possibility that it exists.
Thus, contrary to your assertion, Andrew Sullivan, for example, does not deny the possibility, or ridicule the idea, of a relationship between SSM and marriage in general. In fact, he has argued at some length his belief that SSM is likely to strengthen marriage. What he is ridiculing in the piece you link to is the type of argument that Kurtz and Katz (and now you) are making, i.e., “Trend X started to change around the time event Y occurred, therefore Y caused, or is likely to have caused, X.” Given that there are about a thousand other possible events and trends that are likely to affect trend X, many of which are also likely to have a much greater influence, and none of which you seriously address or control for in your analysis, the whole argument is just ludicrous.
Once more into the breach….
Satire:
1) A literary work in which human vice or folly is attacked through irony, derision, or wit.
2) Irony, sarcasm, or caustic wit used to attack or expose folly, vice, or stupidity.
Satire can be composed of hundreds of thousands of words and can take thousands of hours to complete. Satire is not limited to things that take less than an hour create and less than 500 words to complete. There are satiric novels, plays, songs and even satiric web sites (my favorite was the now defunct Midget Defense League).
I point you to the following quote from this post: “If such an elite educated person says a causal link exists, must it not be so?” I think that, just maybe, that might fit within the bounds of satire.
Jake Squid:
A lengthy and earnest rebuttal of an argument is not satire. There is nothing satirical about Lucia’s response to Katz’s proposed explanations for the trend in out-of-wedlock births.
At the risk of wandering deeper into the pit….
Don P writes, “A lengthy and earnest rebuttal of an argument is not satire.”
We agree there. Kind of.
Lengthy has nothing to do with whether something is satire or not. For proof of that I would like to direct your attention to Dumas’ “The Three Musketeers” which is both lengthy and a satire.
Earnest is important (see “The Importance of Being Earnest”;)). However, if you really believe that, “If such an elite educated person says a causal link exists, must it not be so?” is written in earnest you simply are completely devoid of anything resembling a sense of humor or irony or wit. And for that I am truly, truly sorry. And not a little bit horrified that something so blatantly sarcastic should register as an honest assertion to you.
Now if you want to continue to argue against Lucia’s satirical post as if it were serious…. I guess that that just adds to the whole athmosphere of satire that surrounds this series of posts.
Ladies and gentlemen, on with the show!
(Note: Of course it’s possible that Don P is adding to the humor of the whole thing by pretending to take it seriously, thus duping me. If so, congratulations to Don P on a job well done)
Lucia:
Would you mind terribly going over Dr. Kurtz’s research and finding the exact amount of ink used? That way I figure you could have a running “ink expended” tally in the corner of every post you do with Dr. Kurtz’s amount in one column, representing the goal, and your amount in the other. Y’know, so those of us following along at home can keep track. I want to be there when you take the ink expended total over the top, thereby winning the argument.
Dan J. That is a very good idea!
Clearly, expending ink is they key parameter which proves that one is correct. (Notwithstanding the fact that so many people agree with Julian, despite his parsimony with the stuff. )
It may be slightly difficult to do. I can cut and paste the NRO and standard review articles into a word processor and do a word count. I may have to do a hand count on the pdf file.
Do you think I should count the actual discussion in congress? Or the discussions with news reporters? (He pretty much repeats the same stuff over and over, but the newspapers print it, thus counting ink.)
So far, Dr. Kurtz has published one graph but he has not edited the graph. I published one and later added a lline. Unfortunately, my graph has one less data point, because I couldn’t find the actual birht ratio for 2003. So, blancing out the addition of hte Welfare Reform act and the missing data point, I suppose we have a tie there, right?
Dan J.
I took you up on your suggestion. I cut and pasted my words into a word processor and counted. Assuming no cut and paste errors:
Article 1: 592 words
Article 2: 1894 words (Includes footnotes containing extensive quotes.)
Article 3: 1704 words.
Article 4: 1050 words.
If I include all 4 articles in my tally, I have devoted 5283 words. I believe most people would count slightly modified graph appearing in two articles as 1 graph. However, that is up to the judges.
I should mention, this may over count the number of words I have devoted to addressing issues raised by Katz/Kurtz. After all, Article 4 actually addresses an issue raised by posters at Alas, and K&K. So, some may wish to credit me with only 4188 words.
I then tallied words contained in the Kurtz articles I actually cited. These are:
The affidavit to congress; 4618 words (estimated.)
The feb article: 6050 words’
The may article: 2783 words. (includes a graph)
The june article: 2648
This tallies to 16099 words.
Clearly, even though I fail to account for Kurtz quotes in the April 2004 newspaper article, appearances on talk shows, and in discussion panels, Kurtz leads me by over 10,000 words!.
I must write many more words before I can win by this metric.
I’m thinking that generating a few graphs, which as we all know are worth 1000 words, I can take the lead.
(Note: Currenty, the idea that SSM cannot cause change in the ROI-NMBR is in the lead. So, it looks like I may lose by that metric also. People who believe in cause and effect, please VOTE!)
—
I estimated the words in the pdf file as follows:
1) I picked for sentences at the top of one page, counted the number of words and determined the number of words per sentence.
2) I counted the number or lines in the page, skipping the line with the section heading containng only two words.
3) I multiplied and added to obtain the number of words in the page.
4) I counted the number of full pages. I multiplied. Then I added the number of lines on the final page, and added the estimated number of words on that page.
Well, we’re certainly pulling for you here! Keep it up. I think a few more graphs will help. I think you can also count some of your resposes in the threads. Also, I think the digressions into fluid dynamics are working well for you; not only do they bolster the word count, they also underscore the mathematical consistency of your argument, which is like, way scientific.
Jake Squid:
As I said, there is nothing satirical about Lucia’s rebuttal of Katz’s alternative explanations for the trend in out-of-wedlock births. Her posts on the matter are an earnest argument that Katz’s explanations could not plausibly account for the trend. The problem is that she has become so wrapped up in trying to defeat Katz & co. on their own terms that she has lost sight of the fact that those terms don’t make any sense anyway. She claims that Kurtz and Katz are wrong, but she legitimizes their nonsensical argument by treating the type of correlations they identify as if such correlations actually tell us something meaningful.
Or, at least, that’s the hole she had dug herself into. Her latest series of bizarre entries in her comments boxes suggest that perhaps the penny has finally dropped, and she’s now trying to pretend that she never took Katz’s argument seriously in the first place.
Thanks Dan J.! It is always nice to have a cheering squad. As you say “way scientific”.
It is also nice to hear that fellow bloggers recognize the utility and relevance of statitical fluid dynamics to discussions of this sort. I am hoping that, in future, I can add to the discussion by providing links to pages discussin and instrument calibration. Someone told me section on “model validation” may be relevant is some way.
Lucia:
>I will begin by presenting the case supporting the causal link.
>First, Dr. Kurtz, who has earned a Ph. D.
>in social anthropology from Harvard,
>says there is a causal link between SSM and ROI-NMBR.
>If such an elite educated person says a causal link exists,
>must it not be so?
No. You are subscribing to “false authority” thinking. Just because someone is highly educated doesn’t mean that person’s theories are correct. Plenty of professors are wrong every day, even ones with Ph.Ds.
Please read Propaganda Techniques: False Connections – Testimonial.
>Later on, Rachel Ann noticed I buried Julian’s observation
>deep in the footnotes of this article. Probably suspecting
>my intention to evade Julian’s comments, she posted to
>endorse his point of view. This raised the visibility
>of the idea. Afterwards both Trish Wilson and Don P.
>posted in support of Julian.
That’s a lot of supposition on your part. You’ve committed a logic fallacy, by the way. I was not agreeing with Julian. I was disagreeing with you. One does not automatically follow the other. I also do not think that an unscientific opinion poll will shine more light on the subject.
The problem with your theory is that correlation does not equal causation. That, plus overall unwed birth rates have increased slightly in recent years, which doesn’t say much for the theory that welfare reform has lowered unwed birth rates.
By the way, about that “propaganda techniques” link, Snopes (the Urban Legends web site) calls the same thing “false authority syndrome.” In a nutshell, it means don’t assume someone knows what he or she is talking about solely because that person or entity is an authority figure, has a important position, has many degrees, or is famous. The idea is to remain skeptical because people, being human, are often wrong.
First, dear Amp, I suspected you of nothing untoward in your placement of Julian’s comments; and I still support them. I suspesct that though it is possible that such a link exists, it would be very hard to detct, without taking into account a whole gamit of other possible scenarios.
Second, I took the poll. I think three times. This was not a case of voting early and often but rather that my computer is a disobedient machine and despite my many attempts at discipline (talking to it, threatening, bribery) it continues to behave badly. I think it knows how much I need it and will not really follow through on my threats to banish it to the desert.
Hah. As soon as we have money!!! Alright, that does seem to mean never.
In any case, the computer didn’t seem to be doing anything but staring at me when I clicked on the polls. So I really don’t know how often I voted.
Oops,
Should have said Lucia, but it still holds. I wasn’t
doubting your integrity, I know this debate has been going on for sometime and it is impossible to place everything in the forefront. Footnotes have a purpose and I think you placed the info in the right
space. I just think it highly likely that you are connecting two things which are not causal.
I think that Echinde is right; the cause is quite possibly a Z. In this case, I think it may be that US as a whole has become more conservative, that has lowered the abortion/non-marital pregnancy rate and caused a surge in committed partnering; which for most people would me some sort of formal recognition of the relationship: before G-d or state.
BTW, because I’ve been called on it before, I don’t use endearments in a paternalistic way. I guess that sort of speech is used that way, but I mean it in more of a friendly tone.
To Trish: Thank you for the links to the logical fallacies. I edited the post and added them in locations I believe are appropriate.
To: Rachel Ann,
Don’t worry about overvoting. Evidently, Yahoogroups sorts that out. This is the one advantage of their pesky sysptem where you have to register. They know who votes, and even let you chage your vote up to the time the poll closes. (However, I don’t know who voted or how)
The voting will run 2 weeks from the day I opened the poll.
I believe I will need to modify later articles to indicate that people are agreeing with Echidne. Or possibly, I will simply have to admit that they are in either very direct ways (Don P.) or indirect ways disagreeing with me! ;-)
You’re very welcome, Lucia. Check out Snopes when you have lots of free time. That Urban Legends site is loads of fun.
I think a more effective poll would be to provide the answers and the people taking the poll have to guess what the questions are. ;)
As a stay at home knitter, I have much, much more free time than the average bear! I’ve been enjoying reading Snopes as I compose my next earnest response.
By my count those revisions constitute an additional 52 words, if you count the words added in revision.
Additionally, your two posts in the thread today total out to 157, 158 if the number 2 in 2 weeks is assumed to be read as the word two.
There is an errant p in the word system in your first post, so to avoid gripes from your detractors, it might be safer to just deduct a letter now and have done with it.
You’re getting closer!
There is a short article on this issue in the May/June issue of Off Our Backs, by Karla Mantilla. It mentions Kurtz’s arguments. She also talks about the situation in Norway. I’ve just glanced over the article, but I think she’s arguing that yes, gay marriage will “destroy marriage”, but she’s saying that this is a good thing and could work out to benefit children. Thought I’d mention it due to the recent discussions!
Hi,
Thanks, the mar/april article seems to be online, but I don’t see may/june yet.