When 10% may not actually be 10%

A letter by Chaim Tropper was posted on the blog run by bad4shidduchim. In it he attempted to explain where the "10% of frum women who won't get married" came from. In addition, he attached an article he wrote which deals with the same issue. There are some problems in relying on everything in the letter or the article. I'm going to attempt to set them out, if not in one posting, then in two.

Point #1: When researchers set out to run a study about something they are either trying to prove something they believe, to disprove something they don't believe or to gather information for further study. How successful a study is depends on a number of factors: included in those factors is how the researchers gather their data, what is asked of whom, how terms are defined, are the questions bias-neutral, what is included in the data set reported--and what is left out--and what do the researchers do with the data gathered. Even a slight bias on the part of the researchers, whether intentional or accidental, can skew the results. Even small errors can reduce the practicality of using the results of the study.

Point #2: In the letter to bad4shidduchim Mr. Tropper says: "I took my data from the Avi Chai Study on day schools. By comparing the sizes of different grades I was able to extrapolate the growth rate of the community." He also states: "Briefly the 10% is based on two assumptions. The first is that the community is growing by about 4% a year...leaving compounding out for simplicity."

Point #3: There were two Avi Chai studies on day schools, one for the school year 1998-99 and one for the year 2003-4. Each study dealt with its particular school year; however, the second study also did a comparison of the figures from the first study and the second study. The studies are available to be read at avichai.org.

There are common assumptions used in our communities as to the meaning of "day schools." When we think "day school" we do not think the Lakewood Yeshiva, we do not think the Bobover Yeshiva and we do not think Bais Yaakov schools. For the Avi Chai study "day school" included all schools that offer a dual program of Jewish studies and secular studies, from age four through high school. They broke down "day school" into the following categories: Centrist Orthodox, Chabad, Chassidic, Community, Immigrant/Outreach, Modern Orthodox, Reform, Solomon Schechter, Special Education, Yeshiva. How does this impact on shidduchim?

Point #4: From the content of the article attached to the letter, the concern over the 10% falls mainly over that number as applied to the yeshivish element. So let's look specifically at those numbers across the two studies. For 1998-99: 47,643 students across all grades; For 1003-04: 54,326 across all grades. The change in enrollment is 14.03%. This represents a growth rate, over the five years between the two studies, of 2.80% per year on average. The actual rate of growth fluctuated from year to year. So where did the 4% a year come from? Not from these studies.

Point #5: Mr. Tropper uses the figure of 4% growth each year to explain the 10% of women who won't marry. He says: "Briefly the 10% is based on two assumptions...the second is that boys marry when they are about 3 years older than girls...In a community growing at 4% a year...there are 112 20-year-olds for every 100 23-year-olds (4% a year times 3 years leaving compounding out for simplicity). As more boys than girls are born, this works out to a gap of about 10%." Okay, the study Mr. Tropper says he based his figures on does not show 4% a year as shown above. Over three years Mr. Tropper's figures show a 12% difference; the Avi Chai study showed a difference of 8.4% for the same period. Plugging in 8.4% instead of the 12% used will result in less than a 10% figure for women who won't marry.

But here is the kicker: the Avi Chai studies do not break down any of their categories by gender. That is correct, they do not say how many males or how many females there are on any grade level or in any of the categories. The yeshiva category is defined as having single-sex schools, but the studies do not say how many of the schools are for males only and how many for females only. Where did Mr. Tropper get his information that "more boys than girls are born"? He says that using his 4% figure and the fact that more boys then girls are born will result in a gap of about 10%. Based on what? US government figures show a male/female disparity of approximately 49/51. According to Mr. Tropper's figures, the yeshivish element of the jewish population not only does not follow the general population in that the reverse is true, but that there are 4% more males than females per hundred, per year. Again, based on what study and on what data? Anecdotally we are not seeing an excess of males as opposed to females when we look around.

And let's look at that word "about." Someone posted a comment on here that she went out with a boy who told her he was "about 28." She wanted to know what that meant. I'm curious too. What does "about" mean? Is 9 about 10? Is 8 about 10? We already know that the 4% doesn't hold up and it is the basis for making the "about 10%" statement. If Mr. Tropper's "about 10" statement were actually 9, then figuring in 2.8 instead of 4 would also reduce the 9 quite a bit. Suddenly we are not talking about a 10% group of women.

Point #6: "the second is that boys marry when they are about 3 years older than girls." This is also an important point to consider. Much has been said about how this three year gap is contributing heavily to the 10% group of single women, in that there are proportionately more women born in succeeding years then the year they are looking to marry into. We have all heard that three-year number thrown about, but where did it come from? If generally yeshivish boys are marrying girls three years younger than they are, thus contributing to the 10% overage, how many are specifically not doing so? The Avi Chai study did not cover this at all. So yet again, where are the figures coming from? In the article attached to Mr. Tropper's letter there are some attempts at answering where the three years came from and trying to show that the figure is a true one, but based on no stated figures such as the Avi Chai study provided. And yup, the article is using that 4% growth figure. I'll leave that article for a second posting.

It seems from the lack of data that some of this three-year gap idea is surmising and based on anecdotal evidence. So let's look at some anecdotal evidence. I have a captive audience of male students at school, many of whom are married or "in the parsha." Eleven boys all the same age year--23-- and who are married and who self-select as yeshivish answered the age gap question for me. I asked them to choose one of the following categories: wife is older then me; age difference is less then a year; 12 to 23 month difference; 24 to 35 month difference; 36 to 48 month difference; more then a four year difference. One boy has a wife two months older then he is. 4 boys fell in the 12-23 month difference. 5 boys fell in the 24-35 month difference. One boy fell in the 36-48 category. Of this group, only 5 to 6 fell into the three-year gap category that we all say is what yeshivish boys want. Of course, I would also need to ask when they got married--late nights are death on brain cells.

For someone who professes to hate math I sure am finding myself mired in it lately. But I hope that this posting will at least point out one thing. Just because someone says that something is so does not make it so. Some statistics are born bad; some achieve badness after birth. Rounding is hell when it comes to social figures. If there is a disparity between the number of men and number of women available for shidduchim it may well not be anywhere near 10% and there are other factors besides which might be contributing more. I don't pretend to have all the answers but Mr. Tropper's letter sure doesn't have them all either.

Next up: a look at the article.