class: left, top, title-slide # What’s the matter with polling? ## From
Strength in Numbers: How Polls Work + Why We Need Them
###
G. Elliott Morris
| Oct 18 2022 | Pittsburgh, PA --- <img src="figures/cover.jpg" width="50%" /> --- <img src="figures/tomato.jpeg" width="80%" /> --- # The "soup principle" <img src="figures/tomato.jpeg" width="60%" /> --- class: center, inverse, middle # The first polls --- # "Straw" polls <img src="figures/polls_street.jpg" width="80%" /> --- <img src="figures/digest_poll.jpeg" width="70%" /> --- <img src="figures/digest_1936.jpg" width="60%" /> --- # The first ("scientific") polls ### - Conducted face-to-face -- ### - Used demographic quotas for representativeness - Race, gender, age, geography -- ### - Beat straw polls in accuracy (1936) - By shrinking bias from demographic nonresponse -- --- # The first ("scientific") polls ### - But fell short of true survey science (1948) <img src="figures/dewey_truman.jpeg" width="60%" /> --- # Polls 2.0 ### - SSRC says: area sampling -- <img src="figures/houston.png" width="60%" /> --- # Polls 2.0 ### - SSRC says: area sampling ### - Gallup implements some partisan controls - Strata are groups of precincts by 1948 vote choice -- ### - Use rough quotas within geography -- ### - But, preserve interviewer bias -- --- # Polls 3.0 <img src="figures/phone.jpeg" width="70%" /> -- ### Technological change -> better methods --- # Polls 3.0 ### - 1970s: true random sampling (for people with phones) ### - Response rates above 70-80% ### - Rarer instances of severe nonresponse bias ### - Cheaper to conduct = many news orgs poll (CBS, NYT) --- <img src="figures/aapor.png" width="90%" /> _Source: American Association of Public Opinion Research_ --- # The soup principle: satisfied? <img src="figures/pew_soup.png" width="80%" /> _Source: Pew Research Center_ --- # The soup principle: satisfied? ### 1. RDD polls are representative (at high response) ### 2. Availability of many different surveys allow for extra layer of aggregation to control for choices made by individual researcheers --- class: center, inverse, middle # = perfect polls forever, <br><br> -- # ...right? --- ### Technological change -> worse methods? <img src="figures/pew_response_rate.jpg" width="60%" /> _Source: Pew Research Center_ --- ### Polarized voting -> harder sampling <img src="figures/affpol.png" width="75%" /> _Source: Webster & Abramowitz 2017_ --- .center[ ## But what if the people you sample don't represent the population? ] -- #### - People could be very dissimilar by group, meaning small deviations in sample demographics cause big errors (sampling error) -- #### - Or the people who respond to the poll could be systematically different from the people who don't (response error) -- #### - Or your list of potential respondents could be missing people (coverage error) -- *Polls can also go wrong if they have bad question wording, a fourth type of survey error called "measurement error" --- ## The soup principle in theory <img src="figures/pew_soup.png" width="90%" /> _Source: Pew Research Center_ --- ## The soup principle in practice <img src="figures/minestrone.jpg" width="60%" /> --- class: center, middle # Polls today... -- #### - Declining response rates + Internet = innovations in polling online, but they don't use random sampling -- #### - Traditional RDD and even RBS polls don't have a true random sample (since response rates are too low) -- #### - And because of nonresponse --- ## So, to satisfy the soup principle... ### Pollsters use statistical algorithms to ensure their samples match the population on different demographic targets - Race, age, gender, and region are most common - Can use weighting (raking) modeling (MRP), w various tradeoffs .pull-left[ <img src="figures/raking.jpg" width="100%" /> ] .pull-right[ <img src="figures/mrp.jpg" width="100%" /> ] --- # These adjustments make polls pretty good! <img src="figures/aapor.png" width="75%" /> --- class: center, middle, # But they aren't _representative_, per the theory of sampling -- # ...and in close races, the adjustments aren't enough: --- class: inverse, center, middle # Two examples: --- # 2016: Education weighting <img src="figures/weighting_education.jpg" width="100%" /> --- # 2020: Partisan nonresponse <img src="figures/gq_rs_polls.jpg" width="90%" /> --- # 2020: Partisan nonresponse <img src="figures/gq_rs_polls.jpg" width="40%" /> -- - ### Problem reaching Trump voters overall -- - ### And _within_ demographic groups -- - ### Something you cannot fix with weighting -- - #### Pollsters can adjust for past vote, but the electorate changes, and certain _types_ of voters may not respond to surveys --- class: center middle # So what are we left with? --- # So what are we left with? -- ### 1. Traditional polls that oscillate wildly due to intensive weighting -- ### 2. New "model-based" methods which trade lower variance for higher (potential) bias -- ### 3. Lower response rates increase chance of big misses across firms --- class: center, middle, inverse # Polls (and soup?) in 2022 -- <br> <br> ## A few ways forward: --- # Making polls work again -- ### 1. More weighting variables (NYT) -- ### 2. More online and off-phone data colleciton (SMS, mail) -- ### 3. Mixed samples (private pollsters) -- ### In the pursuit of getting representative (and politically balanced) samples _before and after_ the adjustment stage --- class: center, middle ### In the pursuit of getting representative (and politically balanced) samples _before and after_ the adjustment stage -- ### To satisfy the soup principle --- class: center, middle # What about aggregation? ### Forecasters have a few tricks up our sleeves: --- class: center, middle, inverse # How forecasts work --- # What goes into the model? ### 1. National economic + political fundamentals ### 2. Decompose into state-level priors ### 3. Add the (average of) polls --- # 2. National fundamentals? ### i) Index of economic growth (1940 - 2016) - eight different variables, scaled to measure the standard-deviation from average annual growth ### ii) Presidential approval (1948 - 2016) ### iii) Polarization (1948 - 2016) - measured as the share of swing voters in the electorate, per the ANES --- and interacted with economic growth ### iv) Whether an incumbent is on the ballot --- <img src="figures/fundamentals_economy.png" width="80%" /> --- <img src="figures/fundamental_approval.png" width="80%" /> --- <img src="figures/fundamentals_with_incumbency.png" width="100%" /> --- # 2. The model is a federalist #### i) Train a model to predict the Democratic share of the vote in a state relative to the national vote, 1948-2016 * Variables are: lean in the last election, lean two elections ago, home state effects * state size, conditional on the national vote in the state #### ii) Use the covariates to make predictions for 2020, _conditional on the national fundamentals prediction for every day_ #### ii) Simulate state-level outcomes to extract a mean and standard deviation * Propogates uncertainty both from the LOOCV RMSE of the national model and the state-level model --- class: center, inverse, middle # That's the baseline -- # Now, we add the polls --- # 3. Add the (average of) polls - Just a trend through points... - Can do with any series of packages for R, other statistical languages -- <img src="figures/loess.png" width="50%" /> --- # 3. Add the (average of) polls ### (...but with some fancy extra stuff) ```{Stan mu_b[:,T] = cholesky_ss_cov_mu_b_T * raw_mu_b_T + mu_b_prior; for (i in 1:(T-1)) mu_b[:, T - i] = cholesky_ss_cov_mu_b_walk * raw_mu_b[:, T - i] + mu_b[:, T + 1 - i]; national_mu_b_average = transpose(mu_b) * state_weights; mu_c = raw_mu_c * sigma_c; mu_m = raw_mu_m * sigma_m; mu_pop = raw_mu_pop * sigma_pop; e_bias[1] = raw_e_bias[1] * sigma_e_bias; sigma_rho = sqrt(1-square(rho_e_bias)) * sigma_e_bias; for (t in 2:T) e_bias[t] = mu_e_bias + rho_e_bias * (e_bias[t - 1] - mu_e_bias) + raw_e_bias[t] * sigma_rho; //*** fill pi_democrat for (i in 1:N_state_polls){ logit_pi_democrat_state[i] = mu_b[state[i], day_state[i]] + mu_c[poll_state[i]] + mu_m[poll_mode_state[i]] + mu_pop[poll_pop_state[i]] + unadjusted_state[i] * e_bias[day_state[i]] + raw_measure_noise_state[i] * sigma_measure_noise_state + polling_bias[state[i]]; } ``` --- # 3. Add the (average of) polls -- #### i. Latent state-level vote shares evolve as a random walk over time * "Walks" toward the state-level fundamentals more as we are further out from election day -- #### ii. Polls are observations with measurement error that are debiased on the basis of: * Pollster firm (so-called "house effects") * Poll mode * Poll population * Bias in previous elections -- #### iii. Correcting for partisan non-response * Whether a pollster weights by party registration or past vote * Adjusts for biases that remain AFTER removing the other biases --- # 3. Add the (average of) polls #### Notable improvements from partisan non-response (and other?) issues <img src="figures/states-vs-results.png" width="80%" /> --- class: center, middle # In 2016... -- ## ... But not 2020 -- <img src="figures/2020-economist-histogram.png" width="80%" /> --- # One more lesson: ### 1. Traditional polls that oscillate wildly due to intensive weighting ### 2. New "model-based" methods which trade lower variance for higher (potential) bias ### 3. Lower response rates increase chance of big misses across firms -- ### 4. Aggregation is not a magic bullet --- # 4. Aggregation is not a magic bullet ### What may be more useful than forecasting... --- class: center, middle # Conditional forecasting! --- ## Conditional forecasting: -- .pull-left[ ## 1. Debias polls <img src="figures/conditional_forecasting_one.png" width="80%" /> ] -- .pull-right[ ## 2. Rerun simulations <img src="figures/conditional_forecasting_two.png" width="80%" /> ] --- # 2. Rerun simulations <img src="figures/conditional_forecasting_two.png" width="80%" /> --- # 2. Rerun simulations <img src="figures/conditional_forecasting_two.png" width="50%" /> ### Advantage: leaves readers with a much clearer picture of possibilities for election outcomes _if past patterns of bias aren't predictive of bias_ (2016, 2020) --- class: center, middle, inverse # Further questions: --- # What if that doesn't work? ### 2022 a critical test: does surveys get better or stay the same — or do they get worse? ### What if the DGP remains biased? ### What if the quality of the average poll continues to fall? --- ### Can we trust polls to be precise in close elections? ### If not, what are they good for? --- class: center, middle # How Polls Work <u>and Why We Need Them</u> --- .pull-left[ <img src="figures/cover.jpg" width="90%" /> ] .pull-right[ # Thank you! ### _STENGTH IN NUMBERS_ is Now available. <br> <br> **Website: [gelliottmorris.com](https://www.gelliottmorris.com)** **Twitter: [@gelliottmorris](http://www.twitter.com/gelliottmorris)** ### Questions? ] --- _These slides were made using the `xaringan` package for R. They are available online at https://www.gelliottmorris.com/slides/_