CONDUCTING A COMPREHENSIVE PRECINCT ANALYSIS [Campaigns & Elections]
(Campaigns & Elections Via Acquire Media NewsEdge) Why precint-level modeling is alive and well
here's virtually ^ nothing you can do in 1 a campaign that will s give a better .0/ cost-benefit ratio than a comprehensive i precinct analysis. Precinct analysis looks i at past voting patterns and the dynamic of party affiliation within the context of i neighborhood. It's based on the premise i that people who share similar values, | politically and otherwise, live near one : another. It looks for voting trends, | precinct by precinct, to provide a I geographic location of your core i supporters. And using data from past | elections, it reveals support, swing, and | undervote for similar candidates or i causes. Using this information you may direct your efforts to activate and engage likely support.
This is fundamental to a win: You want to invest effort and resources where they will do the most good.
The ease of locating potential candidate and issue-based supporters through data mining of social media and Internet activity has moved many to declare precinct-level modeling to be dead. However, for the down-ballot campaign, targeting voters based on geographic location is no different than when presidential campaigns identify which states will be battleground territory and which ones will be an automatic win or loss for the nominee. And just as presidential campaigns will not spend i precious resources in states whose outcome is certain, the same goes for the down-ballot campaign. Once cities, communities, and precincts where voters can be persuaded to support your efforts are revealed, you can implement i resources to influence those voters. Precinct analysis is about identifying battleground precincts.
Unlike some campaign activities, a precinct analysis may be conducted in advance of a campaign because it is based on voting history. Take advantage i of this and get it out of the way. It's also cheap. Although election offices charge a nominal fee for past election records, the information you need is public, and relatively easy to obtain. The only exception are counties that do not keep records broken down by party, turnout, or precinct. For such counties, go else; ; to get your data, such as a voter i contact service, your party, or a political j action committee (PAC) that has en: : you and tracks voting history. If it 1 is a voter contact service, be prepared to pay, but your party or an endorsing PAC ; will usually provide some voter contact i information as an in-kind donation to ; your campaign. If your campaign is too ; local for PAC or party support, go to a | candidate who has such support and i offer to include him or her in a precinct I analysis if the person's campaign can get i you what you need.
In addition to voting history from election offices, you may also use U.S. : Census data to further profile your voters with demographic and economic data. The census database will give you demographic information by state, county, and city. This information is a fabulous resource, because data are pre- ; sented separately for each incorporated city, irrespective of size, by neighborhood for larger cities, and by geographic ; area for rural communities. If your area of concern includes one or more small cities, you're in luck.
Whereas polling data will tell you who supports your cause by gender, race, age, income, education, and party affiliation, a precinct analysis will give you an idea of how neighborhoods (precincts) vote on issues and candidates that paral- ; lei your campaign. The analysis also provides a continuum of voter tendencies revealing swing and which party is undervoting their candidate.
There will be times when a precinct analysis is of minimal help: For example, if you are running a partisan race in an area with a huge registration disadvantage and no voter history of a win from anyone in your party. In this example, a precinct analysis will tell you where your voters are lazy-that is, those in your party who tend to be nonvoters-and it may provide some information on swing voters, depending on the quality of the candidates in the previous elections, but that's about it for candidate history information. Nevertheless, knowing where your lazy support lives can : improve your odds come Election Day. i These potential voters can be activated I through lawn sign placement, social I media, direct mail, canvassing, and | phone calling.
Still, unless your opposition is : running his or her campaign from | prison, if the voter registration differ; ; is more than 15 percent, you are : about to launch into an unwinnable race. Indeed, with voters sticking so keenly to party in recent elections, if ; your voter registration disadvantage is : greater than 5 percent, you could be in ; an unwinnable race.
THE SINNERS, THE SAINTS ANOTHESAVEABLES
Basically, there are two kinds of voters: those with their minds made up and those who are undecided. Voters with their minds made up are either for you ("saints") or against you ("sinners"). Those who are undecided or persuadable are the "savables." For precinct analysis purposes, we have to understand the first two groups-the sinners and the saints-but we focus on the third group, the savables.
Sinners will not cast a vote your way, no matter what you say or do so do not spend campaign resources activating them: Don't canvass them, call them, or send direct mail to them. Nothing you say or do will convince these voters to support your candidate or issue. Sinners will vote against you if they vote in your election; they will never cross the aisle. If they do not stay home altogether on Election Day, the best you can hope for is an undervote. Remember, when it comes to sinners, apathy is your friend.
Saints will vote for you, no matter what. They will not cross the aisle, but they can, and do, undervote. That means you do not want to completely ignore these voters, but you also do not want to waste valuable campaign resources on them. *
Savables are registered voters who are neither saints nor sinners, as they do not adhere to party lines. These are the voters you're looking for in a precinct analysis. They tend to be undecided, may be moved by hot-button issues, and often pay little or no attention to politics. Sometimes they are partisans of one party, hiding from their neighbors by registering in another party. And | sometimes they are simply lazy voters that need more prodding.
With precinct analysis, you can i determine where the savables live and then, by further analysis of education, age, and voting history, determine the i likely numbers of those who will actu; ; get out and vote. A precinct analysis ; will give you important information for canvassing and for where to send direct I mail, and even provide a profile for i media, social networking, and phoning.
; CONTEXT OF NEIGHBORHOOD
i Pollsters and political consultants : dismiss geographic context as old j school, but after thirty years of trackI I registration trends both by precinct j and zone in my county and state, I find | context of neighborhood is more relii i than ever for identifying pockets of support. Further, its patterning | dates back to the very founding of the I nation. In his book "1776," David McCullough highlights the adversity the | British troops suffered during the siege of Boston, a city sympathetic to the j revolution, and how anxious they were j to move on to New York, where they would enjoy a warmer welcome. Even then, Americans segregated themselves by political ideology.
But it goes back even further. The genesis of our nation was the coupling ; of religious devotees with those seeking financial opportunity untethered to bloodlines. Not much has changed.
The fracturing within each of the voting blocs on a continuum of time is the essence of a precinct analysis. : Indeed, not all whites vote the same, not all minorities, not all women, not all men, not all eighteento twenty-fiveyear-olds, not all white women between eighteen and twenty-five. What we do today, and have done since getting on ships to cross the Atlantic, is pack up and move next door to people who share core values and think and vote like we do. As Bruce Oppenheimer of Vanderbilt University said, "A lot of this has to do with self-selection. Democrats tend to live next to Democrats. Republicans tend to live next to Republicans."
Precinct analysis shows which neighborhoods and communities are trending where, and more important, it will tell you where the greatest numbers of swing voters live. Once you're in the gold veins, precinct analysis behaves and looks very much like microtargeting because neighborhoods get profiled; that's why it has worked for two hundred years in some form or another.
Knowing the social context of cities within a county is just as important as knowing the social context of neighborhoods within a city. This fine-grained knowledge allows the down-ballot campaign to profile voters within a context of historic voting and registration trends.
For example, when I first began tracking voter registration in 1984, my county was fairly evenly mixed in registration, with a 2 percent countywide Republican registration advantage. Over the course of twenty-five years, this advantage moved from 2 percent to 11 percent and back to 2 percent in 2008. In 1986, the registration difference between Democrats and Republicans in my city was 16 percent; now it's 60 percent. Some communities completely flipped in registration, moving from a 20-point Democratic advantage to a 20-point Republican advantage. Between 2002 and 2012 in one boom town in my county 400 Republicans move in for every 18 Democrats that moved out.
And the same profile that is flipping here is also flipping in small communities in other parts of the state that I've tracked. Although this profile may not hold where you live, in my comer of the world the new Republicans are landed poor, Roosevelt Democrats, blue-collar workers, undereducated people, truck owners, the religious right, survivalist militia, and people deeply suspicious of government. They also have a real resentment for public education and the educated, and endorsements from newspapers are the kiss of death. Where the Democrats are gaining ground is with the educated, the wealthy, smallbusiness owners, Reagan Democrats, secular people, youths, minorities, and unmarried women.
Nearly every sector of my county has become extreme in registration and voting pattem. And when areas become extreme in registration, attitudes and voting will also be extreme, this behavior will filter over to minority registrants as well. For example, Republicans living in heavily registered Democratic areas will be more liberal than Democrats living in heavily registered Republican areas. A Democrat in a very Democratic area may have trouble getting blue enough, just as a Republican in a heavily registered Republican city may have trouble getting red enough.
Polling uses individual traits in conjunction with issue questions to target direct mail and media. For our purposes, individual traits include such things as gender, ethnicity, education, economic standing, and age. Although individual traits are important, exclusively targeting voters using individual traits without social context can lead to missteps. Counties with many small cities, a mix of urban and rural voters, and multiple school districts are the stuff of downballot races. In these races, a strategy that treats all female, nonaffiliated voters-or all those within a certain gender, age, education, or income-as though they were one and the same, irrespective of where they live, is deeply flawed. However, when used in homogeneous populations or primaries, it can be very effective.
So-called independent voters are registered voters who are not affiliated with a political party. In the twenty-seven states and the District of Columbia where voters register by party, the ranks of independents have swelled in the last few decades and now occupy a third of the voter rolls.
Traditionally, a swing vote comes from a Democrat who votes Republican or a Republican who votes Democrat. However, political pundits, the media, and polling firms have elevated the status of independents (nonaffiliated voters, or NAVs) by heralding them as the swing vote and often credit them in postelection analysis for a particular win or loss. Further, in the face of substantive and credible research to the contrary, news and television dramas present independents side by side with Democrats and Republicans, as though they were a political party with a : shared ideology. But this simply is not j true: Non-affiliated voters are closeted I partisans who simply participate in : elections less frequently than their : partisan neighbors. To be successful, a campaign must know how to accurate: : predict their voting behavior.
In partisan races, voters generally must choose between a Democrat and a Republican when casting a vote. Typically, each party maintains a baseline of guaranteed votes, usually about 33 percent. But what about the remaining voters? Can an equally predictable percentage of nonaffiliated and third-party registrants be ascribed to partisan candidates?
After 10 years of observing that NAVs appeared to consistently track party registrants, precinct by precinct, I tested this assertion and developed a simple yet reliable formula to predict voting behavior of non-affiliated and third-party registrants in a given precinct. I've since tested it in counties throughout Oregon and have found it reliable in nine out of ten precincts within one percentage point; the ten percent that fall outside of the formula are where swing voters live.
To predict how non-affiliated and third-party registrants will vote, simply add the democratic registration to the republican registration and divide each by the total. For example you have a precinct with 130 voters; 80 are Democrats, 20 are Republicans, and 30 are non-affiliated voters and third-party registrants. Add the Democrats to the Republicans: 100 and divide each by its total to determine the percent of votes cast that will be owed to each candidate based on registration.
In this example, the Democrat can expect to get (is owed) 80 percent and the Republican is owed or will receive 20 percent of registered voters. Given that Republican and Democratic turnout mirrors each other, a campaign could predict that 80 percent of the registered voters in this precinct will be cast a vote for the Democratic candidate and 20 percent for the Republican candidate. Obviously, in primaries with a bigger fight on one side of the aisle than the other, you must multiply your "votes owed" percentage to the predicted voter turnout, not the registration.
ASCRIBING THE UNDERVOTE
Let's say in the above exampled precinct that the Democrat did not get 80 percent of the votes cast, but rather 79 percent and the Republican did not get 20 percent but rather 17 percent. What happened to the other 4 percent of those voting who did not color in the bubble of either the Democrat or Republican? If there is no third candidate, you will find the missing votes in the undervote column. Typically, voters undervote because they "don't know, don't like, or don't care" about the issue or candidate. But now, with this simple formula, you can determine which party is undervoting which candidate and which neighborhood is sending votes to the other side. In this example Undervotes increase as negative campaigning increases, especially if attacks are deemed as unfair by voters.
FINDING THE SWING
Similarly, if in the above example the Democrat didn't receive the 80 percent he or she was owed but instead received 70 percent and the Republican received 25 percent rather than 20 percent you know that 5 percent of the Democrats swung over and voted for the Republican while another 5 percent undervoted their Democratic candidate. If this occurs election after election, you know it's a precinct that's consistently leaning Republican and will underperform for the Democratic candidate.
A precinct analysis allows a campaign to create a comprehensive blueprint using public information of voting history, the census, and registration shifts. It reveals voter patterns that allows efficient use of campaign resources and offers an affordable and effective campaign strategy for cashstrapped down-ballot races, SB
'This is fundamental to a win: You want to invest effort and resources where they will do the most good."
Catherine Shaw is author of "The Campaign Manager: Running and Winning Local Elections
(c) 2014 Votenet Solutions Inc.
[ Back To TMCnet.com's Homepage ]