Checking a decade of Hawaii's economic forecasts

We compared annual forecasts of Hawaii's economy with the actual results

(page 2 of 3)

Economists acknowledge the limitations of their predictions: Each forecast of the future is simply an expert’s estimate, based on imperfect measurements of the economy at the time the forecast is issued. Nothing more and nothing less.

Council on Revenue Forcasts

COR issues forecasts of the state’s General Fund revenues four times a year, but Hawaii Business focused just on the projections made each March for the coming fiscal year, which starts each July 1. We compared the forecasted increase or decrease with the actual increase or decrease in General Fund revenue for that fiscal year. The average error: 5.54 percentage points.

Click image to enlarge.

Predicting: Visitor Arrivals

Chart shows percentage changes in visitor arrivals. Visitor arrivals are affected by national and global factors, so economists agree they are the hardest local indicator to predict.

Click image to enlarge.

Errors in Visitor Arrivals Forecasts

On average, UHERO’s predictions came closest to reality during the 11 years we surveyed.

Click image to enlarge.

Predicting: Personal Income

Forecasting changes in real personal income after adjusting for inflation

Click image to enlarge.

Errors in Personal-Income Forecasts

DBEDT had the most accurate forecasts on average in this category

Click image to enlarge.

Predicting: Inflation

Forecasted and actual percentage increases in the Honolulu Consumer Price Index. (There is no statewide CPI.)

Click image to enlarge.

Errors in Inflation Forcasts

Click image to enlarge.

* On all error charts, Hawaii Business calculated what we call the average error, whether it was too high or too low. If we had used a traditional measure, the mean error, bad over-forecasts would have cancelled out bad under-forecasts, and created a misleading impression of the forecasts’ accuracy.

Predicting: Job Growth

First Hawaiian Bank and UHERO predict growth or decline in total nonagricultural jobs; for years, DBEDT forecasted total jobs, a slightly different measure

Click image to enlarge.

*The changes in the two job measures, nonagricultural jobs and total jobs, were very close to each other each year: within 0.1 percentage point every year except for 2011, when the two figures were separated by 0.2 percentage point. So we used just a single black bar to show both actual changes in both measures in the chart above.

** Starting with its 2012 forecast, DBEDT began to forecast nonagricultural jobs.

Errors in Job-Growth Forecasts

Click image to enlarge.

Predicting: Unemployment Rate

DBEDT does not forecast Hawaii’s unemployment rate, but UHERO has been doing it for our entire survey period and FHB started doing so with its 2006 forecast.

Click image to enlarge.

Errors in Unemployment-Rate Forecasts

UHERO is most accurate in predicting the next year’s unemployment rate.

Click image to enlarge.

Do you like what you read? Subscribe to Hawaii Business Magazine »