Tuesday, December 21, 2004

Buried Time Series Data on Employment


Here's today's latest Bureau of Labor Statistics report on Regional and State Employment and Unemployment. BLS Regional & State Employment Nov 2004

These reports on based on detailed time series covering many, many variables. Unfortunately, any trends taking place are buried deeply and hidden from view. This 16 page report does not have a single time series chart to show the trends that may or may not be underway. For those who are not faint of heart and love looking at complex tabular data and have lots of time, there are several detailed tables that contain partial time series data for review. There are also a couple of hard to read geographical charts showing current state by state changes in the past 12 months (a implicit time series with only 2 data points).

For those with the time and interest and high speed links and skills, the full time-series can be downloaded, massaged, and analyzed. For the rest of us, we are left with inconclusive text such as the opening summary paragraph:

REGIONAL AND STATE EMPLOYMENT AND UNEMPLOYMENT: NOVEMBER 2004
Regional and state unemployment rates were generally little changed in November. All four regions and 47 states and the District of Columbia recorded changes in their jobless rates of 0.3 percentage point or less from October, the Bureau of Labor Statistics of the U.S. Department of Labor reported today. Over the year, jobless rates declined in all four regions and in 42 states. The national unemployment rate was essentially unchanged over the month at 5.4 percent in November. Nonfarm payroll employment increased in 35 states and the District of Columbia over the month.


This seems to be saying that everything is ok and maybe even good without actually coming out and saying it. The statistics they quote would seem to imply that not much of interest is happening in the regions or the states, that there has been some improvement in the past year (but compared to what), and that employment has grown (but not by how much or compared to anything).

The Economic Policy Institute (EPI) Job Watch takes a look at the same data and comes up with something a lot shorter and a little more user friendly as you can see at: EPI JobWatch: State data and organizations. They also look like they are trying to dig out more interesting facts from the same data, all with a strong point of view that says things really aren't going very well.

Take a look at both of these and see what you think. Who's right? Who's wrong? Who offers more food for thought? Who challenges the reader with more provocative claims?

In neither case, are any of the time series available for easy viewing, so the reader is left depending on the opposing expert writers who prepared these documents or going to a lot of time and trouble trying to actually look at the real data.

Both approaches use what amounts to time-series with exactly two data points and avoid showing us any of the data collected between these two points, or any prior to the earliest points.

Neither leaves me particularly satisfied. Who do I trust when the experts disagree so widely and the data is not readily available?

I would really like to know how our country is doing and what is happening with employment. Are we moving towards a time when we can create enough jobs to keep up with growth in working age population? This is a topic addressed by EPI but completely off the radar in the BLS report.

The EPI main page at: JobWatch does indeed use some time-series to make their points and this is an improvement in my eyes. Still the reader is left mostly dependent on who do they trust when the data reported is so limited.

Finally, take a look at the White House Jobs and Economic Growth: "The State of Our Economy" for yet another way to assess the employment situation. Notice the absence of time series data, and the use of carefully selected two point time series described in words.

Examining cases such as these in near real time, can be instructive in triggering our thinking about how important information can be communicated in a way that gives the reader, viewer, or listener a better chance of judging the facts for themsleves.

2 comments:

Stop & Think said...

Hi Gordon. Thanks. I liked the way you handle the forex data - particularly the multiple time view of the dollar vs the euro and the multiple view of the major currency pairs over a longer time period.

The direction I am aiming for with this blog is to help bring some of the great ideas from the world of finance (as exemplified at your website) and map them over to the other trends that greatly affect the outcome of our lives.

Anonymous said...
This comment has been removed by a blog administrator.