patrick.net

Actually, the numbers show quite the opposite. For something to be statistically independent, like you say, it must follow the formula P(A|B) = P(A), where P(A) is the probability of a recession, P(B) is the probability of a Jewish chairman, P(C) is the probability the chairman is a gentile, and P(A|B) is the probability of a recession given the chairman is Jewish. So, I have said that the probability the chairman is jewish is .47, the probability of a recession is .234, and the probability of a recession while the chairman is jJwish (P(A n B)) is 14.25/95 = .15. So the probability of a recession given the chairman is jewish is shown as P(A|B) = P(A n B) / P(B), is .15/.47 = .32. Remember the probability of a recession is .234. They are not equal, therefore they are not statiscally independent. 32% likelihood of a recession given the chairman is jewish.

Similarly, the probability of a recession given the chairman is a gentile, P(A|C) is as shown: P(A n C) = 8/95 = .084, P(C) = .53. So .084/.53 is .16. 16% likelihood of a recession given the chairman is a gentile.

Looks like twice as likely to me.

Further, since there have been no presidents that are not christians, an analysis will yield nothing. 100% of recessions occur during christian presidencies.

If you'd like to suggest something real to research and show correlation between, I would be interested honestly. Ivy league educated presidents and recessions? Brown eyed presidents and recessions? Whatever.

Why don't you spend your time trying to figure out what they are doing right then? You said that Jewish people dominate those other areas, personally, I don't know. My guess would be that Hindus dominate biological sciences, asians dominate technology, etc. But I haven't done the research.
Personally I would think that once you come out acting like a jerk and then get shown the facts like you did, you should probably go away with your tail between your legs.
I'll play along. What are the economic policies that you would like to show correlation to recessions? Monetary expansion? Keynesianism? Deregulation? Let's start with Keynesianism. You find the years is which this was the dominant monetary policy and I'll do the correlation analysis.

@David--

The data set is small enough that I guarantee that the Jewish question is statistically insignificant. You can't prove anything either way. Recessions are often determined by economic decisions made several years in the past--quite possibly by a previous Fed Chairman.

True, the data set is way too small. I think much of the mess is due to Greenspan, with Bernanke's hand forced in many places by Greenspan's earlier actions. At least forced by the orthodox economic response, which sees the cure for excessive debt as yet more debt.

About ethnicity and economics: I think it's true that Jews make better economic choices than the majority, but they have the advantage of growing up with relatives and a community that knows exactly what those better choices are. Each group has a kind of cultural momentum from its past. Poke around construction workers and short-order cooks, and you'll find they don't have a clear idea of the admission process for medical or law school, or the expected salaries once you get out. Admission seems impossible to them personally, chatting with their co-workers and family doesn't help, and so their kids don't make the better choices early enough either.

I could see this in my own family. My dad was the last of 8 Irish kids and the only one to go to college. But he had to fight the momentum, which was towards just getting any job out of high school. His family told him college was a waste of time and money. My mother, raised Jewish, stressed college and graduate school for us and my father agreed. Now my brother is a lawyer, and I got into medical school myself, but chickened out and did not go.

Not that you can't escape your cultural milieu, but it's much harder than just going along with it. They will mock you for being different.

david1 says

They are not equal, therefore they are not statiscally independent.

You should not confuse probability with statistical analysis. Probability of a Jewish chairman is not 0.47, it is between .38 and .57 with 95% confidence and during recession is between 0.158 and 0.24. With these in your conditional probability P(B|A) = (0.16,0.62). No significance. Apparently you attended the Stats100 class on conditional probability but slept through the part about statistics.

For your next assignment poster, please analyze the Jew ratio of Congress or at least this panel. Not that I put any stock in your argument at all, it just seems an obvious omission.

I do find it startling how the Peter Principle seems to guarantee ALL of these morons keep their jobs.

Bernanke may have PLAIN problems like he's either an idiot or a liar when he said things like "subprime is contained". Or the dozen other things he's either been WRONG on or LYING about. However the people who appointed him and will probably re-appoint him, what's their excuse? So I have to include Congress as either incompetents or criminals.

I predict outcome will be "stay the course" which is the Bush variant of "you shouldn't change horses in mid-stream" which seems to apply even if the horse is the one that drug you into the water in the first place. Only this stupidity usually assumes that the horse is half-way across a stream and just need to hang on and you'll be out safe and dry. What if your idiot horse is instead swimming out to sea?

totally,

what confidence interval are you using? Isn't this a binomial confidence interval (jewish or not, recession or not). If it is, then the confidence interval can be approximated by the normal distribution by CLT as shown:

P +- z(1-.95/2) * sqrt{p(1-p)/n}
n is number of years, p is estimated proportion, z(1-.95/2) is the z score corresponding to the 95th percentile of the standard normal distribution. so, p for jewish chairman, .47+- .063(sqrt(.47*.53/95). I get a confidence interval of .467 to .473. With such a small confidence intervals, I don't think it will change much.

In short, we sample each of the 95 years (large sample, n>30) and have two binomials, jewish or not, recession or not.

I used:
http://www.causascientia.org/math_stat/ProportionCI.html

This is a confidence interval for proportion not for mean.

Nomograph says

on the whole, Jews tend to have above-average intelligence, a hard work ethic, and place a huge premium on education. It is almost impossible find a Jewish construction worker or short-order cook.

Are you telling me, you don't know any Jewish fuckups?

I find that hard to believe. Not every Son can be like Eliot, I don't care what race or religion you are.

David1 - Repeat after me: "Correlation is not causation."

Thanks totally, very helpful. What if I had chosen to sample both binomials on a daily, instead of yearly, basis? For that matter, every second? As n becomes large, the range of the confidence interval shrinks. The proportions really would not change very much based on how precisely it is measured in this case. I think that is why the confidence interval is large for low sample sizes, as low sample sizes increase the chances of sampling errors.
lyoungblood,
I don't think I ever said that simply because the two may be correlated, Jewish folk have caused recessions. I simply read Bernanke's biography, which lead me to read about other chairmen. I thought it odd that such a high proportion of them have been Jewish, leading me to get into the numbers a little further. I did not mean to imply anything, I was just thinking out loud, reporting what I had found and thought it was interesting.
I still think it is pretty amazing that such a high relative percentage of chairmen of the Fed have been Jewish. Even if EVERY Jew in the country was in the top 5% income bracket, AND there was perfect 100% correlation between being in the top 5% income and eventually becoming chairmen, it would only be expected that a Jew would be chairman 44% of the time. What we have experience thus far exceeds that. I find that amazing.

He can't be chairman one second then not a chairman the next. the units should probably be average term length or something ... but that would be complicated.

Sure he can, he could die, or resign, etc. Also we can be in a recession one second and not the next. Even though a recession is reported on a quarterly basis, there is one moment in time where a economy changes from increasing to decreasing. That is not the precision in which this data was gathered, but you do see the problem with confidence intervals here.

No ... I totally buy the confidence interval. What I am saying is that you can't switch chairmen every second. That is the independent event this interval relies on. But you can switch chairmen every term and that should be the unit. But anyways ... there is not enough data. Wait for a hundred more chairmen and recessions and we will see what it looks like :)

Well I hope he gets a terrible finger lashing from Senate committees, before they unanimously reaffirm his post.

In the unlikely event that he isn't confirmed... who would replace him, and could they do any better (or worse?)