Monday, October 08, 2012

More about polls

Most of the mainstream polls have had Obama in the lead for the last several weeks. However, one significant poll is one in which Romney leads -- the enthusiasm measurement.

The Battleground Tracking poll shows that Romney has a 13-point lead among voters who are "extremely likely" to vote.  What's more, Romney's leads are significant among elderly and white voters, who have traditionally been high turnout groups.  Meanwhile, Obama leads among young voters, blacks and latinos, who -- except for 2008 -- are not traditionally energized to vote, and their enthusiasm is down significantly from their 2008 highs. 

So, although polls which ask random Americans or even registered voters may show Obama with a big lead, those likely to vote favor Romney.  However, as I've said before, the only poll that really matters is the one on November 6.

Saturday, October 06, 2012

About those unemployment numbers

The government's official unemployment numbers came out Friday, and they showed an unemployment rate of 7.8%, falling below 8% for the first time in 44 months. This obviously gave Obama a boost and a chance to discuss something other than his no-show at the debate.

The unemployment numbers caught everyone by surprise. Thursday's reports had predicted a rise to 8.2%, due to several economic factors. Obama's Labor Secretary was forced to come out and defend the numbers, falling back on the "we're all professionals" defense. Convincing, except for a few anomalies.

A summary which accompanies the announcement of unemployment figures is the "Summary of Recent Trends." It's not as boring as it sounds, because of charts like this one:


A bit of explanation is in order.  The official unemployment numbers are based on the "Household Survey," based on census figures. Using that figure, the economy apparently added 873,000 jobs in September. Another metric surveyed is the "Payroll Survey" of businesses, which literally speaks to businesses to see how many jobs they've added. 

Over the entire last year, the difference between the number of jobs reported between the two is 850,000 jobs. (Nothing sinister there, they use different measurements.) However, 759,000 jobs of that difference occurred in the single month of September. 

September was quite a month. October's "revisions" should be interesting.