Question: How did you determine the stimulus plan would be a success?
Steven Castellano: So at the Moody's Challenge basically you wake up you get your computer set up with a group of five kids and at 7 am you download a document and this is year was the stimulus package and how effective will it be. And I know the first hour I would say was the most difficult because I don't really think we've got a lot done. You're kind of just doing a lot of research and you're wondering how we're going to do this, how are you going to be able to analyze the stimulus when like economists are doing this 24/7 and they haven't reach upon conclusion and you know, you just kind of say what assumptions are we going to make and how we are going follow through with it and that first hour is probably the most nerve-wracking in that sense because you don't really know your direction and you start to plot a course that may not be successful.
So what we ultimately did was we look at how much money was being spent in each sector, if it's education, or an infrastructure and then we look at how many jobs are generally created per dollar in that sector by looking at like an average of few if you graph the total amount of revenue versus the total amount of jobs, you'll get kind of get an approximate for the rate of how money jobs per dollar you'll get and then in addition to that then we did like an infinite serious calculation in which you look at like the marginal propensity it consume and that's basically how much money a person will spend per dollar and you multiply that infinitely because as one person gets the money in the new job and they'll spend it and then I could promote more job for another sector.
So that's kind of the first analysis we did to see how many jobs will be created and we determine that there would be about 3.1 million jobs created using that analysis. We also found that in terms of like bank for your buck the most jobs per dollar it wasn't education sector because of our, as I said the total jobs for total revenue sort of analysis we did and taxes were actually, were actually the least like bank for your buck but because they are the largest part of the stimulus they did create the most jobs.
So you know, we weren't necessarily sure if we agree with the breakdown of the stimulus. We would say maybe less tax cuts and more spending on education but overall we found that it would probably create 3.1 million jobs in that sense and in addition to that, then we ask to analyze unemployment. So we look at previous trends in unemployment with the 1930s recession I think 60s recession and 1980s recession and we found that first statistically significant decreased unemployment occurred 3 years after the first statistically significant increase on federal spending and all of three of these instances. So we kind of made a linear regression because I have the least degrees of freedom for federal spending is the per cent of the GDP versus the unemployment rate. When we did this we noticed that it quickly became negative unemployment at the federal spending that Obama is proposing because it is unprecedently high.
So then we, what we did was we saw that there is Okun's law and it relates federal spending and unemployment it was like a natural algorithmic relationship, so we then apply the natural algorithmic regression toward data set and found that there was actually a stronger fit in with the linear regression and we are able to approximate that. Since it's a three year offset factor between the unemployment rate and federal spending percent of GDP is that, this federal spending in three years or reduce the unemployment rate to about 6.2 percent versus without the stimulus it would be more in the order of 9 percent so that was like our second major analysis. We also set conditions for if the 2nd stimulus will be needed. What benchmarks do we expect as far as the unemployment reaching 6 percent in three years and if we need a second stimulus, we mention the breakdown and we said that like food stamps and education spending would, would probably the most beneficial so that was our analysis and we have five of us working 14 hours straight in a crazy hectic environment but we definitely produce a pretty cool paper if you consider that a brief time spend and just like the passion that went into it and produce.
Recorded on: May 8, 2009