Poll | Date | Sample | MoE | Republicans (R) | Democrats (D) | Spread |
---|---|---|---|---|---|---|
PPD Average | 5/10 – 5/26 | — | — | 37.0 | 38.5 | Democrats +1.5 |
Quinnipiac | 5/19 – 5/26 | 1711 RV | 2.4 | 36 | 39 | Democrats +3 |
Rasmussen Reports | 5/10 – 5/14 | 2500 LV | 2.0 | 38 | 38 | Tie |
Rasmussen Reports | 5/3 – 5/7 | 2500 LV | 2.0 | 38 | 38 | Tie |
Rasmussen Reports | 4/26 – 4/30 | 2500 LV | 2.0 | 40 | 38 | Republicans +2 |
Rasmussen Reports | 4/19 – 4/23 | 2500 LV | 2.0 | 39 | 38 | Republicans +1 |
Quinnipiac | 4/16 – 4/21 | 1353 RV | 2.7 | 40 | 37 | Republicans +3 |
Rasmussen Reports | 4/12 – 4/16 | 2500 LV | 2.0 | 39 | 39 | Tie |
Rasmussen Reports | 4/5 – 4/9 | 2500 LV | 2.0 | 40 | 37 | Republicans +3 |
Rasmussen Reports | 3/29 – 4/2 | 2500 LV | 2.0 | 40 | 36 | Republicans +4 |
Rasmussen Reports | 3/22 – 3/26 | 2500 LV | 2.0 | 39 | 38 | Republicans +1 |
Rasmussen Reports | 3/15 – 3/19 | 2500 LV | 2.0 | 39 | 39 | Tie |
Rasmussen Reports | 3/8 – 3/12 | 2500 LV | 2.0 | 39 | 38 | Republicans +1 |
Rasmussen Reports | 3/1 – 3/5 | 2500 LV | 2.0 | 41 | 36 | Republicans +5 |
Rasmussen Reports | 2/23 – 3/1 | 2800 LV | 2.0 | 39 | 39 | Tie |
Rasmussen Reports | 2/16 – 2/22 | 2800 LV | 2.0 | 38 | 39 | Democrats +1 |
Rasmussen Reports | 2/9 – 2/15 | 2800 LV | 2.0 | 39 | 38 | Republicans +1 |
Rasmussen Reports | 2/2 – 2/8 | 2800 LV | 2.0 | 40 | 39 | Republicans +1 |
Rasmussen Reports | 1/26 – 2/1 | 2800 LV | 2.0 | 38 | 40 | Democrats +2 |
Rasmussen Reports | 1/19 – 1/25 | 2800 LV | 2.0 | 37 | 40 | Democrats +3 |
Rasmussen Reports | 1/12 – 1/18 | 2800 LV | 2.0 | 38 | 39 | Democrats +1 |
Rasmussen Reports | 1/5 – 1/11 | 2800 LV | 2.0 | 38 | 38 | Tie |
Rasmussen Reports | 12/29 – 1/4 | 2800 LV | 2.0 | 38 | 40 | Democrats +2 |
Rasmussen Reports | 12/22 – 12/28 | 3500 LV | 2.0 | 39 | 39 | Tie |
Rasmussen Reports | 12/15 – 12/21 | 3500 LV | 2.0 | 40 | 39 | Republicans +1 |
Rasmussen Reports | 12/8 – 12/14 | 3500 LV | 2.0 | 40 | 37 | Republicans +3 |
Rasmussen Reports | 12/1 – 12/7 | 3500 LV | 2.0 | 39 | 40 | Democrats +1 |
Rasmussen Reports | 11/24 – 11/30 | 3500 LV | 2.0 | 41 | 40 | Republicans +1 |
Rasmussen Reports | 11/17 – 11/23 | 3500 LV | 2.0 | 43 | 39 | Republicans +4 |
Rasmussen Reports | 11/10 – 11/16 | 3500 LV | 2.0 | 41 | 41 | Tie |
Rasmussen Reports | 11/3 – 11/9 | 3500 LV | 2.0 | 42 | 41 | Republicans +1 |
Latest generic congressional ballot polls, or generic ballot polls, and PPD average of weekly tracking and snap-shot polling surveys.
People’s Pundit Daily includes both Democrat and Republican partisan polling companies, that is, up until a month prior to the election when we begin weighting pollsters’ based upon accuracy and past performance due to pollsters attempting to influence the generic congressional ballot average. Read 2014 House midterm election predictions below the table.
Juxtaposed to other polling aggregators, People’s Pundit Daily computes averages on a tracking basis. In other words, unlike many others, our polling average does not include stale data, which is why we are far more accurate. Stale polling data misses momentum, recent and relative trends, and even rare but possible extreme swings in the electorate.
Brief Analysis Of 2014 Midterm Election And PPD Model Predictions
People’s Pundit Daily gives the edge in the 2014 House midterm elections to the Republican Party for a number of reasons.
First, despite what advantage in the generic congressional ballot the Democratic Party may have — if any — come Election Day, House districts are drawn in a manner that favors Republicans. Democrats would need a very statistically substantial lead on pre-election generic ballot surveys — in fact, somewhere in the neighborhood of 12 to 14 points — in order to have any chance of gaining the 17 House seats needed for them to regain control of the chamber.
Because the generic congressional ballot is a national survey, it often cannot reflect what may be occurring on the political landscape in the battleground regions of the country. When the generic congressional ballot shows a lead of 2 – 5 points in favor of Democrats, it is more than likely that the sentiment in heavily-populated urban, concentrated liberal areas is obfuscating or understating support for the Republican Party in battleground or conservative-leaning districts.
For instance, some pollsters found Democrats leading the generic congressional ballot right up until Election Day in 2010, but that was a wave election for the Republican Party, who won a total of 242 seats with a net gain of 63. Let’s take a look at all 17 post-World War II midterm election results, compare them to generic congressional ballot results, and draw some real conclusions all other variables excluded. It is worth noting a few observations that standout before getting to the actual model’s predictive value.
Until the summer and early fall of the actual midterm election year, the results of the generic congressional ballot have little predictive value. Further, the better Republicans are performing on the survey in early September, the more actual seats the Republican Party is likely to gain, which may skew early predictions in favor of the Democratic Party. Also, historically Republicans pick up more seats on a point-for-point basis. In other words, on average, a 10-point advantage for Democrats yields just a 12-seat swing, while an identical lead for Republicans yields a 23-seat swing, on average.
[table id=11 /]
From the table above, we get a visual picture both of how much the generic congressional ballot understates Republican support, and the daunting task facing Democrats. Considering Republican redistricting after their 2010 gains, it is entirely possible that this model actually understates their support further, though their potential pick-up opportunities — as I’ll explain further shortly — will most likely limit any chance to properly gauge that in 2014.
As we have hammered over and over, presidential job approval is a statistically significant variable when gauging midterm election outcomes since 1980. In 2010, though the Democrats aren’t as exposed in the House this year, President Obama’s approval rating was roughly 45 percent. In fact, because there are fewer opportunities in 2014 and Republican pickup opportunities are likely to be limited, we might expect less damage this time around even though Obama’s approval rating may just be lower.
Since 1912, when the House expanded to 435 seats, the president’s party has lost an average of 29 House seats in the “six year itch” midterm election. However, Democratic losses are likely to be limited in 2014 due to the fact that Democrats are only defending 201 seats this year, and it is all about the particular districts.