Of course I can. The change in seasonally adjusted non-parm payroll employment was +195,000 (Table B-1
) Using Table 3
for AE (all employee) 1 month change from preliminary estimate, we see that the standard error is 55,726
Using 90% confidence, (1.645 standard errors) the margin of error is +/-91,669 so with 90% confidence the actual change was between +103,331 and +286,669 Zero is not included, therefore the change is statistically significantly different from zero.
CPS data is a little more complicated. Going to Table D1 in the Household Data appendix
, starting on Page 195 for explanations, we see that for Full time employees the "a" parameter is -.0000164, the "b" parameter is "3095.55" and the factor for 1-month consecutive change is 1.11
The average of May and June 2013 for full time workers is (116,238,000 + 115,998,000)/2 = 116,118,000 The change is -240,000 as seen in Table A9
Plugging it into the formula we see that the Standard Error is 1.11*(-.0000164*116,118,000^2+3,095.55*116,118,000)^.5 = 412,827
At the 90% confidence interval, that's 1.645*412,827 = 679,100 making the actual change in full time workers between -919,100 and +678,860 which does contain zero and is therefore not statistically significantly different from zero.