JT and I are devoting a portion of my weekly appearance to the Grand Forks Flood of 1997, and the business and economic consequences of the event. The consequences of some events are best understood with the perspective of time, and natural disasters are clearly this type of event. For today’s post I chose to look at unemployment, but in a slightly different way.
The unemployment rate is a conventional measure for this, but I chose to look at the year-over-year percentage change in unemployment. Why this different look? I think it does a better job relaying the complexity of the response of a local economy after experiencing a natural disaster.
Here is what the traditional unemployment rate (unadjusted) looks like:
There is a significant increase in the unemployment rate in May of 1997, immediately following the flood in April of 1997, so no surprises there. There is a decrease and then possibly a slight increase, though from this graph I think it difficult to completely determine what happens with the dynamics of the unemployment rate after that.
Instead, I prefer to use the percentage change, year over year, for the level of unemployment rate. A different type of unemployment calculation, yes, but again I think the richness of the graph makes it worthwhile.
This graph exhibits the same spike up immediately following the flood. Completely unsurprising and not very interesting. The other graph showed that too. The incredible dip in unemployment afterwards, also speaks to circumstances after the flood. This graph displays the dynamics of the market adjustment after the flood in a way we do not see in the graph of unemployment rate. The spike up is followed by a dramatic decline as people both return to work and leave the local market, migrating out. There is then another increase as markets search out a new equilibrium, though the variability indicates how long a process that actually was.
I think this a far superior representation of what labor market adjustment and transition was like after the flood in 1997. It also gives us a jumping off point for further investigation. What exactly explains the drop? How much is due to outmigration? And how much is due to a resurgence in economic activity to repair flood damage?
