How Good Are The Projections?
by Silence Dogood

The fourth issue of the HuskyData Newsletter, “a regular newsletter dedicated to sharing data and information about SCSU and our students,” was released on Tuesday, November 4, 2014 and focused on enrollment projections.

The first figure shows the Fall 2014 NEF (New Entering Freshmen) Forecast Over Time. The original projection on February 5th was for an increase of 122 students over the Fall’13 enrollment which is an increase of 7.1%. That’s pretty amazing growth if it turned out to be true. By the February 19th projection, the possible increase dropped to 77 students yielding a healthy growth of 4.5%. That’s still pretty amazing growth.

By the March 5th projection, the number was reduced again to the value from Fall’13 for a net growth of 0%. From that point, the projections vary ±2% above and below the actual 30th day NEF enrollment.

So, for a month from February 5th to March 5th things looked pretty good or at least not too bad. The actual 30th Day NEF number of 1,683 represents a decline of 1.2% from the Fall of 2013 to Fall 2014. It’s still a decline, but at least less of a decline than the prior year. The following Figure shows the NEF Headcount from Fall’08 through Fall’14.

The headcount enrollment decline from Fall’08 to Fall’14 is 718 students, which amounts to a decline of 29.9%! Clearly, the percentage decline is staggering. Hopefully, the trend will not continue too far into the future. SCSU cannot survive as the second largest MnSCU institution, with much more declining enrollment. Believe it or not, this fall SCSU has only 33 more NEF than Winona State University!

The second graph appearing in the HuskyData Newsletter shows Forecasts of Fiscal Year 2015 FYE Enrollment at various points in time. The February 5th projection was for an FYE enrollment of 12,125. This represents a projected decline of 256 FYE for a 2.1% decline from the FY14 enrollment.

Unfortunately, the February 5th projection significantly overestimated the actual enrollment. “After adjusting for the unanticipated decline in summer enrollment,” the current projection is for an enrollment of 11,798 FYE, which is a loss of 583 FYE and a decrease of 4.7% from the FY14 enrollment. Looking at the plot, in May, the projection dropped by approximately 250 FYE but the summer enrollment was only down 100 FYE corresponding to a decline of nearly 10% in summer enrollment. Clearly, something more significant than the decline in summer enrollment affected the FYE projection in May.

If you look at the first enrollment projection in the HuskyData Newsletter compared to this final enrollment projection (11,798), the projection decreased by 327 FYE and by itself corresponds to a net decline from the prior year of 2.6%. Essentially, the enrollment decline for the latest projection is more than twice the original projected decline of 2.1% (an error in estimation of 120% error).

President Potter has repeatedly praised the Data Analytics Workgroup for the accuracy of their enrollment projections. If the budget for FY15 was built using the first enrollment projection, the big question is how much of a financial effect does being off by 327 FYE represent. Assuming each FYE produces $11,500 in revenue, 327 FYE represents $3,760,000. Considering the current projected budget deficit is $9,542,000, this ‘error’ represents 39% of the projected budget deficit. Clearly, whether or not the enrollment projection was right on target or off by 327 FYE, the budget is going to have to be cut by a very big number. However, one might like to know the bad news earlier so plans could be made to anticipate the shortfall rather than have to react to it.

Looking at the FYE projections for FY15, the latest data point shows the FYE number from November 2014. At this point, summer enrollment is already in the bank. Fall semester is more than half way through so the final fall numbers will only change by a small amount. Students have already started registering for Spring semester. Making a projection for Spring semester and ultimately the total FYE enrollment for the year at this point should be pretty accurate. Quite simply the majority of the FYE enrollment for the year is already known and there is pretty good historical data about the percentage of students returning in spring semester from fall semester. It’s kind of like counting the number of people in a movie theater at the beginning of the movie. Certainly, some people come in late but the change in the total number is very small.

If the enrollment projection of 11,798 comes to pass, the enrollment decrease will fall within the range 4-5% as revised in May (the announced March projection was for a 3.2% decline). While we can all argue about projections, I believe the final FYE FY’15 number will be closer to 11,719, which would correspond to a decline of 5.3%. We’ll see the actual 10th day enrollment numbers shortly after the beginning of classes in January. Either way, a decline of 4.7% or 5.3%, when coupled where the university started from its peak in FY10 (15,096), the enrollment is down over five years either 21.8% or 22.4%. This seems like a distinction without a difference to me. What is not a distinction without a difference is the over estimation of enrollment that then results in budgeting expenditures greater than the actual revenue generated. The effect of the over estimation of enrollment has resulted in the creation of the need to cut expenditures after a majority of the budget has been committed.

One can only hope that the error in estimation does not continue and the Data Analytics group learns from it’s mistakes and that the budgeting process will be based on more reliable estimates of enrollment and revenue. Another $3.7 million surprise for next year on top of this year’s $9.5 million dollar surprise surely will create major financial problems for SCSU.

One Response to “HuskyData 4”

  • “Another $3.7 million surprise for next year on top of this year’s $9.5 million dollar surprise surely will create major financial problems for SCSU.”

    I see Chancellor Rosenstone is too busy collecting no confidence votes to provide any meaningful oversight to SCSU or any other MnSCU campus. Oh wait…he never did before so why start now?

Leave a Reply