I was looking at labor force data again and thinking about the consecutive months of increase in North Dakota. I went back to 2001 for the data and found that North Dakota enjoyed a streak of 53 consecutive months labor force increase from August of 2009 to December of 2013. Now this is just a simple increase, not a percentage change or any type of scaling, so an increase of one represents an increase, though it may seem trivial. During the oil boom the increase was typically larger, with an average increase of 777.5 workers per month, but even if a few months were low numbers it was still an impressive streak.

On the face of it that number may not seem like a great deal. However, the relative concentration of those workers would be significant and in many cases in areas that were not set up to handle an increase of hundreds of new workers per month. It is also the case that there are times that the labor force increase in a place like Williston was greater than the state increase, meaning other places in the state lost labor force while Williston gained. Anyway, that is not the point of this post so we will come back to that later if needed.

So I then got to thinking about this streak and how it stacked up against other streaks in other states. I went back as far as 2001. This was not for any defined statistical reason necessarily, rather it seemed to make sense as a reasonable length of time, and was after the technology bubble of 1999. I will look at sub periods in a further post but thought I could see the scale of this streak in this expansive time frame. I was surprised at how far down the list it ended up.

The Bakken labor force streak (as I decided to name it) ranked 18th best over this time frame across the country, taking only the best streak from each state. First was New Mexico with a streak of 117 months. 117 months is almost ten years of consecutive increases! The streak started with the beginning of my time period so it could be longer, but it ended in October of 2008, a likely as a consequence of the Financial Crisis. The second longest streak, 116 months for Nevada, also started at the beginning of the time period and ended one month earlier. It seems likely the end of the streak was due to similar events. In fact the other two streaks of over 100 months (Utah (109), South Dakota (101)), also ended around the time of the Financial Crisis, and so did the fifth longest streak (Washington (90)).

Clearly then it seems the Financial Crisis interrupted streaks for many state economies, which is not really a surprise. Though it should be noted this was not universal. In addition, there is enough time in this sample for there to be multiple different streaks, and in many cases extended streaks that may be quite long, or two long streaks broken by only one or two months of minor decline. I will address that in a later post too. The longest active streak right now appears to go to Minnesota at 75 weeks.

I thought I would provide a visual for this however and decided that rather than time series or other graphics I would generate a simple binary representation. The graph is of decrease or increase in labor force and is for all fifty states, and Washington, D.C. over the entire sample period.

The lack of pattern here is actually interesting. There are no solid yellow lines across the entire time span, and there are no solid purple lines across the entire set of states. Clearly there are some horizontals with more yellow than purple, and there are some verticals with more a dominant color as well. Again, this does not say much of anything about the magnitudes of change, which I will address in a later post.

It surprised me that the Bakken labor force streak was not one of the longer streaks in the country. Within North Dakota it was of monumental importance in the state, potentially altering the trajectory of the economy for a generation. Whether the streaks were sustained is another interesting question that can be addressed, potentially, but does get into cyclical considerations as well that can be difficult to evaluate across states at times.