During Mississippi Economic Council’s annual fall “hobnob” at the Jackson coliseum, our state leaders waxed eloquently about our state’s economic health.

Lt. Gov. Tate Reeves said Mississippi had the 17th fastest growing economy in 2012. MEC president Blake Wilson said, “Our trend line is moving in the right direction in numerous economic indexes.” Delbert Hosemann said, “Mississippi has achieved an annualized growth rate of 4.6 percent in the last six months.” Gov. Phil Bryant mentioned several magazine rankings touting the economic progress of Mississippi.

There are some positive signs: The state tax commission statistics on retail sales show an increase of 3.65 percent for the year ending in June 2013. Even better, the first three months’ retail sales are up 6.39 percent. Government tax collections are ahead of projections and reflect the retail sales.

Unfortunately, the most important indicator of all — job growth — is showing mixed results. Of the two major surveys done by the federal Bureau of Labor Statistics (BLS), one shows an increase of 18,200 jobs and the other shows a decrease of 22,318 jobs. The difference between the two BLS surveys is a whopping 40,518 jobs — 3.4 percent of the total jobs in Mississippi.

Mark Twain said there were lies, damn lies and statistics. How can the two major employment surveys show the exact opposite?

The residential survey is derived by asking workers if they have a job or are looking for a job. The establishment survey is derived by asking business establishments how many workers they are employing.

I posed this question to Darrin Webb, our state economist. He emailed me the following:

“One surveys households, the other establishments. Generally the establishment series is considered the more reliable. It also allows you to look at employment by sector. The household survey is used to calculate unemployment rates and that is about it. Based on the establishment series, employment is up over 18k jobs. On average for 2013, employment is up 1.7% over 2012 – that is looking at average employment January through August. That is the fastest growth since 1999. Hope this helps.”

So there we have it. Two surveys. Two opposite results. Webb trusts the establishment survey as do most economists, he says. That’s good news.

However, the residential survey, which shows significant losses in jobs, is the survey used to calculate the official federal unemployment rate. It is not easy to simply dismiss it.

“It would be better if they were both showing the same trend,” Webb told me.

Ironically, during the depths of the great recession the roles of the surveys were reversed. The establishment survey showed job losses and the residential survey showed job growth. The theory was that people who were losing establishment jobs were becoming self-employed.

There are many other theories about why the surveys don’t agree. If someone takes a second job, this extra job would show up on the establishment survey but not on the residential survey.

If someone has become discouraged from looking for a job, retires, goes on disability or decides to stay home to raise children, these would all have a big effect on the residential survey, but would not show up on the establishment survey.

Indeed, the residential survey shows the Mississippi labor force decreasing by 43,390 people from August 2012 to August 2013. That has caused the state unemployment rate to drop from 9.3 percent to 8.5 percent even though 22,318 jobs were lost during the same time period.

If workers were coming in from out of state, they would show up on the establishment survey but not the residential survey. The huge number of Alabama workers at the Kemper power plant could be one reason for the discrepancy, Webb pointed out.

A sudden decrease in self-employment would show up on the residential survey but would not show up on the establishment survey.

On top of all that, both surveys often undergo some major revisions over time. Webb said that revisions and discrepancies are particularly major during these particularly tumultuous economic times. During periods of stability, the surveys tend to agree.

All of these surveys, by the way, can be found by Googling “bureau of labor statistics” and going to their Web site.

So which is it? Have we just had a terrible year of job loss or a great year of job growth? Establishments are reporting job growth but individual Mississippians are reporting job losses and dramatic labor force declines. Mark Twain had it right. Stay tuned.