Here is a response I received from someone affiliated with project Nasi. He sent me a graph but I wasn't able to post it the way he sent, I assume you'll get the idea though.
"His data is messed up.
1) Look at his year-over-year growth rate, which goes negative for a few years
2) He shows that there are 20% more boys than girls born every year. Ludicrous.
If you have 20% more boys than girls each year you need a 4-year or 5-year age-gap to even things out. These are numbers that cannot be trusted.
Here is his data in the first three columns and the simple extrapolation using these numbers.
Grade girls boys more boys than girls % more boys than girls YoY growth girls YoY % growth girls YoY growth boys YoY % growth girls
Actually, their numbers aren't that far off from mine. I've also seen one or two years of negative growth - although the overall trend is positive growth. And, particularly using my larger dataset, I'm also seeing significantly more boys than girls: in Ocean County the surplus is 10.6% and in Brooklyn it's 8.5. I can't explain such numbers yet - although I suspect some of it is possibly due to a large number of Queens/5 Towns families who send their girls to schools in Brooklyn.
I'm very curious about where they got their data. If you can have them get in touch with me through admin@bootstrap-it.com I'd appreciate it.
This fellow was pointing out that according to your numbers there are 20% more boys than girls, which sounds ludicrous.
Also, I just edited my first comment to include quotations, so you see that most of the comment is him talking. It wasn't clear originally. My apologies.
Oh. I see. Sorry about that. Those are certainly good questions.
So first of all, those are the actual numbers from the reported enrollment for (nearly) all frum schools in NY, NJ, and PA. These aren't the result of assumption-heavy extrapolation from general population data or small-sample surveys - which can, of course, both be valuable in some limited contexts. This is much harder data.
The only way I can imagine it's possible to reject "my" numbers is if you assume that the mosdos were lying when they reported their enrollment to the government. I suppose that's theoretically possible, but for a lot of reasons it's very unlikely.
I think it's worthwhile drilling into the the numbers to try to better understand them rather than just rejecting them. I'm working on that with my larger dataset right now and hope to publish more about it in the future.
Secondly, the large numbers of boys can be at least partially explained when you break the data down by county. As I wrote above, a lot of the discrepancy is possibly due to a large number of Queens/5 Towns families who send their girls to schools in Brooklyn.
It might also be worthwhile getting more people to look at the data. That way, it'll be more likely that we'll identify strange details - like, for instance, a large school that's not represented. It could also help us understand anomalies like the "disappearance" of around 40% of boys as soon as they hit 9th grade.
The Ami article looked at actual births, which I think is a better place to start, and they have it at a 106.7 to 100 ratio, nothing near 20%.
An issue with Ami's data is that they were only counting Lakewood, whereas in the past 5+ years many of the Lakewood community moved to Jackson, Toms River etc. So they didn't get an accurate year on year population growth.
The Ami article data was probably pretty good, although you do have to remember that at least 15-20% of that population isn't Jewish, which does skew things.
The truth is that my data is nowhere near 20% either. The problem was that the smaller dataset I used for the article didn't include Queens or Nassau County - where there are far more girls than boys (because, I'm assuming, a lot of the girls go to school in Brooklyn). If you look at the ratio in each county individually, the numbers are far closer.
That's actually true. I've corrected it now, though. Although I've since assembled a much larger dataset that includes Queens and Monsey and covers around 85,000 students in total. In this one the male surplus over each of the years between boys' grades 5-8 is 488, 238, 409, and 395 above and beyond the number of girls who are *four* years younger. And the male surpluses for the same years in Lakewood (Ocean County) are 108, 8, 92, and 59. Whichever way you slice it, there's no statistically significant gap that could cause a marriage crisis.
Looks like excellent research! That website is a great discovery! Would you be willing to share your excel file? I understand if you don't want to do that.
I'm (almost) always willing to share my data. I've got some issues with a larger dataset I created subsequently, so I'd love another pair of eyes on it. Send me a direct email by replying to the original message.
Here is a response I received from someone affiliated with project Nasi. He sent me a graph but I wasn't able to post it the way he sent, I assume you'll get the idea though.
"His data is messed up.
1) Look at his year-over-year growth rate, which goes negative for a few years
2) He shows that there are 20% more boys than girls born every year. Ludicrous.
If you have 20% more boys than girls each year you need a 4-year or 5-year age-gap to even things out. These are numbers that cannot be trusted.
Here is his data in the first three columns and the simple extrapolation using these numbers.
Grade girls boys more boys than girls % more boys than girls YoY growth girls YoY % growth girls YoY growth boys YoY % growth girls
1 3205 3928 723 22.6% 2 0.06% 8 0.20%
2 3203 3920 717 22.4% 157 5.15% 117 3.08%
3 3046 3803 757 24.9% 52 1.74% 430 12.75%
4 2994 3373 379 12.7% 305 11.34% -66 -1.92%
5 2689 3439 750 27.9% -44 -1.61% 176 5.39%
6 2733 3263 530 19.4% -140 -4.87% -97 -2.89%
7 2873 3360 487 17.0% 109 3.94% 66 2.00%
8 2764 3294 530 19.2%
No point in examining this erroneous data."
Actually, their numbers aren't that far off from mine. I've also seen one or two years of negative growth - although the overall trend is positive growth. And, particularly using my larger dataset, I'm also seeing significantly more boys than girls: in Ocean County the surplus is 10.6% and in Brooklyn it's 8.5. I can't explain such numbers yet - although I suspect some of it is possibly due to a large number of Queens/5 Towns families who send their girls to schools in Brooklyn.
I'm very curious about where they got their data. If you can have them get in touch with me through admin@bootstrap-it.com I'd appreciate it.
There was an issue with the graph I uploaded.
These numbers are your numbers not theirs.
This fellow was pointing out that according to your numbers there are 20% more boys than girls, which sounds ludicrous.
Also, I just edited my first comment to include quotations, so you see that most of the comment is him talking. It wasn't clear originally. My apologies.
Oh. I see. Sorry about that. Those are certainly good questions.
So first of all, those are the actual numbers from the reported enrollment for (nearly) all frum schools in NY, NJ, and PA. These aren't the result of assumption-heavy extrapolation from general population data or small-sample surveys - which can, of course, both be valuable in some limited contexts. This is much harder data.
The only way I can imagine it's possible to reject "my" numbers is if you assume that the mosdos were lying when they reported their enrollment to the government. I suppose that's theoretically possible, but for a lot of reasons it's very unlikely.
I think it's worthwhile drilling into the the numbers to try to better understand them rather than just rejecting them. I'm working on that with my larger dataset right now and hope to publish more about it in the future.
Secondly, the large numbers of boys can be at least partially explained when you break the data down by county. As I wrote above, a lot of the discrepancy is possibly due to a large number of Queens/5 Towns families who send their girls to schools in Brooklyn.
It might also be worthwhile getting more people to look at the data. That way, it'll be more likely that we'll identify strange details - like, for instance, a large school that's not represented. It could also help us understand anomalies like the "disappearance" of around 40% of boys as soon as they hit 9th grade.
Please feel free to pass that along.
The Ami article looked at actual births, which I think is a better place to start, and they have it at a 106.7 to 100 ratio, nothing near 20%.
An issue with Ami's data is that they were only counting Lakewood, whereas in the past 5+ years many of the Lakewood community moved to Jackson, Toms River etc. So they didn't get an accurate year on year population growth.
The Ami article data was probably pretty good, although you do have to remember that at least 15-20% of that population isn't Jewish, which does skew things.
The truth is that my data is nowhere near 20% either. The problem was that the smaller dataset I used for the article didn't include Queens or Nassau County - where there are far more girls than boys (because, I'm assuming, a lot of the girls go to school in Brooklyn). If you look at the ratio in each county individually, the numbers are far closer.
How exactly are you calculating the "surplus" on your chart?
Seems like you're calculating a 3 year age gap, not 4 years. Correct?
That's actually true. I've corrected it now, though. Although I've since assembled a much larger dataset that includes Queens and Monsey and covers around 85,000 students in total. In this one the male surplus over each of the years between boys' grades 5-8 is 488, 238, 409, and 395 above and beyond the number of girls who are *four* years younger. And the male surpluses for the same years in Lakewood (Ocean County) are 108, 8, 92, and 59. Whichever way you slice it, there's no statistically significant gap that could cause a marriage crisis.
Looks like excellent research! That website is a great discovery! Would you be willing to share your excel file? I understand if you don't want to do that.
I'm (almost) always willing to share my data. I've got some issues with a larger dataset I created subsequently, so I'd love another pair of eyes on it. Send me a direct email by replying to the original message.