all comments are subject to moderation. your comment will post once approved.
At the February 24th meeting of SANDAG, the San Diego Association of Governments, the Board of Directors comprised of the 18 jurisdictions of San Diego County, heard an explanation of the bad forecast used to anticipate revenue generated from Proposition A, the 1/2 cent sales tax increases which failed on the November 2016 ballot.
Ray Major, SANDAG's head of technology services and chief economist, explained through a series of power point slides the problems which he has uncovered in the formulas used by the agency to forecast revenues and other economic indicators. Only having been in the post for a little over a year, Major described finding a problem with the Retail Sales Tax estimates, which generated the emails the Voice of San Diego published in their story.
The Board took this opportunity to express their concerns and request an independent investigation into the details surrounding who knew what and when and why the Board had not been made aware of the mistake much sooner.
While this is a matter of great importance, the Board glossed over issues raised which are much more serious indicators of the problems inside SANDAG.
During his presentation, Major disclosed SANDAG uses and maintains close to 20 terabytes of information in a serious of databases. After this issue, SANDAG has decided to employ a Database Administrator. It is concerning at the least to think with all the data and databases being used, SANDAG would employ someone to maintain the data's integrity and the database integrity. Hopefully, moving forward, after this mistake, they will.
Also, Major showed a slide during his presentation which showed a flow chart of how exactly the Retail Sales Tax estimate figure is calculated. The flow chart shows approximately 15 steps and calculations which happen before the estimate is reached. Up to this point, SANDAG had never mapped the flow, according to Major. The agency had never previously given this much thought.
At its crux, the investigation seem to point to a flawed process which skewed the data being put into the projection. SANDAG currently purchases national economic data from Moody's. Moody's provides data to SANDAG, according to Major, broken into 391 differenct economic sectors. SANDAG only uses 50 sectors and requires a manual condensing of the Moody data.
While Major was hopefully oversimplifying the description of the mistake, he said it basically came to choosing the wrong cell in an Excel spreadsheet. Instead of choosing the correct cell, they chose a neighboring cell which was more enthusiastic in its forecasting. Essentially, it was human error which caused the problem.
The ramifications of this error are astounding. Not just in terms of an understatment of projected revenue for a sales tax increase. But, in terms of how bad forecasting has potentially affected all 18 member cities.
All cities in the State of California are required to update their General Plans - the document which regulates the growth of their cities - on a regular basis. Some elements of the document are required to be updated more frequently. The most important elements of General Plans - the Land Use and Housing Elements - determine the availability of new housing stock in cities and where the cities will put it. This information is based on forecasting estimates provided by SANDAG.
Cities have been spending millions of dollars on these plans which could all essentially be invalidated, or at least seriously flawed, because the data which was used to project estimates was flawed from the source.
SANDAG has a much more serious problem than bad publicity; the agency must get its data and economic forecasting house in order. Until then, the reputation of the Board of Directors will continue to be sullied. The Board must take direct and immediate action to correct this mistake. They began the process by approving an independent investigation of the issue. They must do more.
Photo credits: SANDAG logo - wikipedia.com, slides from SANDAG presentation
All opinions expressed are those of Don Greene and not necessarily his employer.