Posted - Filed under Household Survey, Release Observations.

The unemployment rate continues to fall, but for what must be considered a disappointing reason: The percentage of the population who are either in work or are regarded as seeking work (the Labor Force Participation Rate) continues to fall, while the percentage of the population who have a job (the Employment/Population Ratio) has not risen.

Employment Rato Unemployment Rate and the Participation Rate


In the two charts above. we plot the unemployment rate against each of those two determinants.

In the last four years the share of the population which has had a job (the Employment/Population Ratio) has gone sideways. Of itself, this would cause the unemployment rate to hold level too.

But, the participation rate has dropped steadily, tending to remove relatively more people from out of the ranks of the unemployed than of the employed. A disproportionate number of unemployed have been removed from the rate equation, and the unemployment rate has thus fallen.

People join and leave the Labor Force for various reasons: they leave school, they go back to school, they retire or they give up looking for work. Age-related retirement’s effect was expected to be on the increase as the bulge of the baby-boom reaches retirement. But, a look at the age-related statistics would suggest that this is not the only nor the most significant reason for the drop in the Participation Rate.

For more charts and tables at finer levels of detail, please visit our Household Survey Dataset Page.

December Revisions
Although the seasonal adjustment process revises all prior months with every monthly release, the BLS does not publish the revisions. Except with each year’s December survey. The reason for this is that the revisions came as part of a new seasonal adjustment process many years ago, and the BLS decided to keep their old no-revisions publishing procedure in place “to avoid confusion”.

U3 RevsionsThe chart at right compares today’s (“December”) consistent time-series with the previously reported sequence of monthly (and hence, strictly, not comparable) data points published up until November’s survey.

The differences are not dramatic – as the BLS claim, in defense of their stance again publishing monthly revisions – but we have seen in the past year or two how a month to month change of 0.1% can mean all the difference to their readership.