The May release came out earlier this week and was a real whopper. It was quite high and well outside the 95% confidence interval of the forecast from last month. We need another month, or months, of data to determine the impacts of the tax increase though. This number could be high in anticipation of the higher rate, or it could be pent up demand shifted to March due to bad weather in January or February. Easter also occurred in March this year and, as my forecasting class saw, that increased sales tax collections in Grand Forks in the past so it could be that situation again. It could also be related to tax cuts at the federal level though I am a bit skeptical that it would just start showing up in spending data for March. Like I said though we need to see where it is at over the next few months before determining the longer term trajectory. Here is the updated forecast.
Alright, we got the update from the city and it was not pleasant. Year-to-date collections are down more than 10%, and cities preferred measure (though we do not know why it is preferred) is down 6.87%. The forecast was for $1.35 million in collections and it came in at $1.02 million. So my pessimistic forecast was still too rosy. Based on this the updated forecast for May is revised down to $1.3 million.
Over the past few weeks the conversation about sales tax has been pretty consistent and fairly intense. That is probably as it should be given the overall importance of revenue generation across the state right now. What have I done so far? I think the best thing to date may be the discussion surrounding the rolling twelve month series used by the city. Rather than accuracy it is more about stability. Eleven of twelve months in the series are the same from one month to the next so of course the series is stable. The flaws of this approach can be seen when you look at the recent decline. A good month is totally outweighed by the totality of the previous eleven. It also begs a statistical question: should all these months be given equal weight while in the series and then just disappear? That seems highly suspect.
Recent posts clearly place me in the camp of “critic” for the sales tax data from Grand Forks. Criticism can be important because it encourages us to make things better, to improve upon the inadequate. With that in mind I offer up some alternatives that might be better than the current system. If we need to remove some of the “ups-and-downs” of the data, for whatever reason, let’s try to keep the some of the features of the data that keep us in the realm of informing policy. This first graph is for a long period of time and may be a bit difficult to see.
Last week I argued (here) that the city’s use of the rolling twelve month total was not really about accuracy as much as stability. Essentially the two concepts are conflated in the presentation. I still fail to see the virtues of knowing the total sales over the last twelve months, and how it better informs policy decisions. I present the following graphs to demonstrate my points about stability. Here is the monthly collection amount (recall there is approximately a two month lag between sales tax collections and the reported amount).