The lecture I gave at the Society for Nonlinear Dynamics and Econometrics Symposium in San Francisco two months ago on regional business cycles (earlier summarized here and
here) is now available via video streaming (courtesy of Bruce Mizrach and the SNDE). Note you have to click through the slides yourself as the lecture proceeds, and it seems to work with Internet Explorer but not Firefox.
Too cool – thanks for posting the link!
The presentation cut out on me, i.e. reset to t=0 about 15 minutes in. But it’s fascinating to see see the process in visual terms.
Thanks!
That was great! The question about realtime data right at the end asked a question I was interested in. Have you had a chance to look further into that?
I love the slide show with the video. So long as the slides are well labeled (yours were) it is easy to follow.
You speak very well. You seemed both casual and prepared at the same time.
Nice talk- the “animated maps” were indeed impressive.
I am not an economist but had undergraduate statistics classes many years ago.
However, I wonder if there is any bias in the data due to “closed data”: the fact that the unemployment rate is similar to concentration data in chemistry, by definition i.e. can’t exceed 100%.
Besides, in the USA, the unemployment rate rarely exceeds 10%. Thus, it still can be treated as a statistically totally independent and “free” variable.
To illustrate this, consider a scenario: if a state reports more than 10% unemployment, this alone might put it into the “late recovery” type of state. (Just becuase there are fewer people left that can join the workforce… Well, kind of).
Are there any states that had data points with 20 or 30% unemployment? Must such data points receive special treatment?