I have already answered the problem except the last part of it. Since it is hard to submit the data table with the question I have put my answered below ( which is the right answer ) to provide you with the data you would need to do the explanation.

Economists often make forecasts of future conditions. Consider the US unemployment rate for June 2015 as predicted in October 2014 as part of a survey of economists, as shown in Table 8.6.4.

Mean = 5.59173913

Standard deviation = 0.22097063

The average forecast of June unemployment rate for the 46 economists surveyed is 5.59173913, the standard deviation of the predictions is 0.22097063. This is small standard deviation tells us that the spread of the data is small. The mean, median and mode of the data are all nearly equal, and %99.7 of the forecasts falls between 4.92882724 to 6.25465102 which includes all data but one forecast.

The standard error of the forecasts is 0.032580342, this tells us that the true mean of the population lies somewhere between 5.559 and 5.624.

= 5.59173/ 5.3 = 0.29173913

The prediction is way off. It is 0.29 less than the prediction.

d.How many standard errors away from the sample average is the actual outcome (5.3%)? Would you ordinarily be surprised by such an extreme difference?

= 0.29173913 / 0.032580342 = 8.954452687

The outcome is 8.954452687 standard errors away from the prediction. In a normal, situation, we would be alarmed by the difference, so we must wonder if our sample was bad or too small, or did something happen in those eight months? It's a prediction, that's all.

Here is the question; Explain why the forecast error (average forecast minus actual outcome) need not be approximately equal in size to the standard error. Do this by identifying the population mean and showing that the actual outcome is not the same object.