Public Policy Institute of California
Graded against the actual result across 20 races (from 39 polls, through 2022).
Head-to-head vs VotePredictor Elections
The fair, apples-to-apples test: on the 12 races Public Policy Institute of California actually polled, how its final poll's margin compared to what VotePredictor Elections predicted for those same races.
| Model | Avg miss (pts) | Called right |
|---|---|---|
| Public Policy Institute of California | 4.32 | 100% |
| VotePredictor Elections | 2.39 | 100% |
VotePredictor Elections aggregates all the pollsters, so it's expected to beat any single one on margin — that's the value of averaging. The honest comparison among forecasters is on the combined board.
Every race (18)
Each race Public Policy Institute of California polled, scored on its final poll — the call right before the vote — against the actual Dem−Rep result. Click a race for its full detail.
| Race | Their call | Result | Miss | Call |
|---|---|---|---|---|
| 2022 CA Governor | D+19.0 | D+18.4 | 0.6 | ✓ |
| 2020 CA President | D+26.0 | D+29.2 | 3.2 | ✓ |
| 2018 CA Governor | D+11.0 | D+23.9 | 12.9 | ✓ |
| 2016 CA President | D+26.0 | D+30.1 | 4.1 | ✓ |
| 2014 CA Governor | D+16.0 | D+19.9 | 3.9 | ✓ |
| 2012 CA President | D+12.0 | D+23.1 | 11.1 | ✓ |
| 2010 CA Governor | D+8.0 | D+12.9 | 4.9 | ✓ |
| 2010 CA Senate | D+5.0 | D+10.0 | 5.0 | ✓ |
| 2008 CA President | D+23.0 | D+24.1 | 1.1 | ✓ |
| 2006 CA Governor | R+18.0 | R+17.0 | 1.0 | ✓ |
| 2004 CA President | D+12.0 | D+9.9 | 2.1 | ✓ |
| 2004 CA Senate | D+18.0 | D+19.9 | 1.9 | ✓ |
| 2003 CA Governor | D+2.0 | R+17.1 | 19.1 | ✗ |
| 2002 CA Governor | D+10.0 | D+4.9 | 5.1 | ✓ |
| 2000 CA President | D+5.0 | D+11.8 | 6.8 | ✓ |
| 2000 CA Senate | D+17.0 | D+19.3 | 2.3 | ✓ |
| 1998 CA Governor | D+8.0 | D+19.6 | 11.6 | ✓ |
| 1998 CA Senate | D+3.0 | D+10.0 | 7.0 | ✓ |
Accuracy by time to election
Lower is better. Time to election runs right (election week) to left (~2 months out).
By the numbers
| Time to election | Polls | Avg miss | vs field | Called right |
|---|---|---|---|---|
| 1–3 wk | 8 | 3.87 | -1.20 | 100% |
| 3–6 wk | 14 | 7.04 | +1.30 | 86% |
| 6–9 wk | 17 | 6.60 | +0.45 | 100% |
vs field is this pollster's average miss minus all pollsters' at the same lead time — green beats the field, redtrails it. Our historical polls reach ~2 months out; earlier polling isn't in the record.
Track record by cycle — getting better?
| Year | Polls | Avg miss | Lean (house effect) |
|---|---|---|---|
| 1998 | 4 | 9.3 | R+9.3 |
| 2000 | 4 | 3.5 | R+3.5 |
| 2004 | 4 | 2.0 | D+0.1 |
| 2010 | 4 | 6.7 | R+6.7 |
| 2016 | 4 | 8.9 | R+9.1 |
| 2018 | 4 | 8.8 | R+12.4 |
Do we credit a pollster for fixing its bias? Each cycle, the model re-estimates every pollster's lean from all its earlier polls (walk-forward) and subtracts it before using the poll. We tested weighting recent cycles more — it doesn't help: a pollster's lean in one cycle barely predicts the next (correlation 0.28), so the swings above are mostly noise, and averaging over more history beats chasing the latest cycle. The all-time estimate we use came out within ~0.5% of the best option.