The Texas Tribune “Inside Intelligence” campaign predictions are a reflection of Austin groupthink, which is driven by risk-avoidance, not analysis.
Last cycle the same Texas Tribune Inside Intelligence believed, with a 91% to 8% margin, that David Dewhurst would beat Ted Cruz. Warren Chisum was picked to beat Christi Craddick 58% to 37%. Jeff Wentworth was picked to retain his seat by 85% of the Tribune’s Insiders, Elizabeth Ames Jones got 13%, and Donna Campbell, the eventual winner, was picked by 2% of the insiders.
Each of these were the safe opinions. Predictive quality is not really the point. Austin insiders are expected to help build a sense of inevitability for establishment candidates. If you pick the non-establishment choice and he wins, you are seen as smart, but a little reckless. There is risk in this approach.
By contrast, if you pick the establishment choice and the other candidate wins, it is easy to turn your lack of vision and faith in the challenger into a compliment. Just say, “Wow, you did great! Everyone thought you were a long shot! I doubt you’ll be underestimated again!” There is no risk in this approach. It is a crack in the wall for capitol city cockroaches to sit in.
Now the Austin groupthink, reflected in the newest Texas Triune Inside Intelligence, is picking Dan Branch over Ken Paxton by 40% to 28%, Todd Staples over Dan Patrick by 27% to 26%, and John Cornyn over Louis Gohmert 88% to 10%.
Austin insiders are bureaucratic-type survivors, not reliable predictors of campaign outcomes, making these insider opinion polls virtually worthless.
–April 24, 2012 Texas Tribune Inside Intelligence
–August 28, 2013 Texas Tribune Inside Intelligence