Emails, Joshua Wright, director of Marketing and PR, EMSI, Sept. 3-4, 2014

2:06 p.m.

Sept. 3, 2014

The calculation, based on my quick look at the spreadsheet, does hold up. If the candidate is saying it’s 57% and our data is showing 58% (or 57%), the claim looks accurate to me. It’s a simple calculation of job change for occupations that have median hourly earnings below the living wage. (I would mention that the data only includes traditional employees of firms, not the self-employed. That’s not a huge deal because this data lines up pretty closely with the payroll data released by government agencies.)

As far as whether 57% is a big deal, I’m not sure I’m equipped to answer that in an educated way. It would probably be best to compare Austin’s share with other similarly sized metros to see if it’s dramatically higher.

Best,

Josh

Joshua Wright

Director of Marketing and PR

EMSI

On September 3, 2014 at 12:39:23 PM, Selby, Gardner (CMG-Austin) (wgselby@statesman.com) wrote:

Would anything hugely substantive have changed in the months since we did that story?

Chewing on this some more, I wondered too if these results are as dire as the candidate said, in part because the median wage isn’t the same as the average wage. Thoughts?

Another facet: Just because many of the jobs added in the Austin area had median wages higher than the living wage for an adult with a child -- the indicator Zehr singled out for his story -- it’s still the case that many of the individuals holding those jobs were paid less than the living wage; after all, we’re looking at median pay. This make sense?

Conversely, though, isn’t it incorrect to indicate that each of the jobs the candidate referred to didn’t pay a living wage?

Another aspect: There are lower living wages listed on the MIT website including, for both counties, $18.51 for two adults and a child; $14.88 for two adults only; and $9.43 for one adult alone. It stands to reason that greater shares of the jobs gained in the Austin area have median wages higher than these lower living wage rates. According to the spreadsheet, 40,560 of the 92,620 net jobs gained from 2009 to 2013, or 44 percent, had median hourly wages less than $14.88; 15,099 of the net jobs gained, 16 percent, had median hourly wages less than $9.43. Is this meaningful when we’re talking about families and wages in the jobs gained?

?

g.

9:34 a.m.

Sept. 4, 2014

Nothing crazy has changed. The only thing I can think of is that the BLS released the 2013 OES survey in May and we’re about to incorporate it in our next quarterly data release. So the data Dan wrote about is based on 2012 wages, but 2013 wage information is out. As I mentioned in my last email, though, our occupation data is benchmarked to our industry data, largely because there are such wild fluctuations from year-to-year with the OES Survey. (The BLS strongly suggests not to look at trends with OES for this reason.)

Median wage is a statistically more relevant figure than average wage because it’s influenced by wild fluctuations at the extreme. Median is simply the 50th percentile wage (half make more, half make less). You are correct, though — some individuals in those occupations were paid less than the living wage. But some were paid more, too.

On September 4, 2014 at 1:49:27 PM, Selby, Gardner (CMG-Austin) (wgselby@statesman.com) wrote:

Where did the estimated median wage for each profession listed come from?

4:28 p.m.

Our occupation wage data (median, average, etc.) comes from the Bureau of Labor Statistics (its OES survey), and it’s benchmarked to our high-quality industry data, which also comes from a variety of public agencies like the BLS and Census Bureau.