Dr. Philip Castillo
Polling is both an art and a science. As an art, it involves personally interacting with people and as a science, the information obtained from that interaction are revealed in probabilistic forecasts. An abundance of polls were undertaken for the just concluded midterm elections in the USA. In following both the polling and election night returns, some comparisons to Belize became obvious.
America has nearly 260 million voters broken down into 51states, including its capital and over 2,200 counties. Belize has 212,000 voters broken down into six districts and 31constituencies. A poll starts when the pollster obtains a sample frame. This is the list of all voters who can potentially vote. In the USA, this list has phone numbers, thereby making potential voters reachable via that means of communication. Belize’s electoral office does not provide phone numbers. In any event, using only the phone to undertake polls introduces errors called non-coverage and non response errors. Non coverage since there are voters who do have not have phone numbers or listed phone numbers. Non response occurs when voters do not answer their phones or consent to be polled. Pollsters have ways of addressing these errors.
From the sampling frame, pollsters draw a random sample, around 1,000 in the USA. Am not aware of any national political polls in Belize, but our statistical agency, the Statistical Institute of Belize generally uses a sample size of around 3,000 for its national surveys. The sample is a small scale replica of the population and pollsters do their best to ensure that all the characteristics of the population are captured in the sample. However, there will always be a difference between the larger voting population and the smaller sample drawn from that population. Pollsters use a statistic called margin of error to account for this difference and accurately measure this potential error.
Incumbency is supposed to offer advantages. The incumbent government has the resources of the government to use in its campaigning. The incumbent can also engage in pre-election chicanery to consolidate its advantages. In the USA, there was widespread gerrymandering undertaken by the Republicans . In Belize, the known chicanery also occurs pre elections when during the months of July and August, there is the voter transfer period. Voters are paid to transfer to a new constituency where they nominally reside at a “safe house.” In reality, they have absolutely no affiliation with that constituency, but will show up on election day. Gerrymandering didn’t work for the Republicans. The jury is still out whether voters transfers have any meaningful influence on final results in any constituency in Belize.
Money has an outsized influence in both countries. In the USA, the world’s richest man Elon Musk reportedly spend near a half billion US dollars on the elections. He lost bigly. The two recent internal elections in Belize offer contrasting findings. Purely from observation in a country that has no campaign finance laws, in the PUP mayoral race in Belize City, the candidate whose campaign appears to have spent the most money, won. But in the PUP Corozal Town mayoral election, the reverse seemed to have occurred. What gives?
An obvious similarity may be incumbency. In both Belize City and Corozal Town, the incumbent won. Small clarification, the incumbent Mayor in Belize City was term-limited but he threw his full support behind his Deputy. Perhaps incumbency does offer advantages since incumbency provides visibility or voters may genuinely be satisfied with the status quo. But perhaps there may have been other variables at play. Perhaps the wining PUP candidates in Belize City and Corozal ran genuinely superiors campaigns. What about candidate quality concerns? Did that factor in voting behaviour?
But back to money. It can’t be separated from politics. The consensus is that it’s better to have lots of money than to not have. But perhaps more importantly than the amount of money is how a campaign spends the money that it has. Former President Barack Obama used volunteers to knock on doors and operate phone banks. Voters were asked if they could bring at least five other persons to the polls? If you can’t bring five, then at least three? But do bring along someone. Sometimes not having money encourages creative thinking.
In the USA, early voting is normal and in some states, it is possible to vote weeks before the actual election day. Not so in Belize, where all voting is done on the actual election day. In Belize, proxy voting occurs and this provides yet another opportunity where voters can be bribed. Both countries have laws against bribing voters. In both countries, such laws are ignored. Bacon houses have now become a terminology in Belizean politics.
In the USA, voting by mail is routine and this facilitates participation from America’s population spread globally. Not so in Belize, despite its known and acknowledged diaspora population. To vote they must come home on that day.
Polling predicts, but more than that, it also shapes campaigns, ascertain candidate characteristics that attract or repel voters. Polling provides information that campaigns can use to sharpen their messages, reallocate resources, give themselves an electoral advantage and otherwise dictate campaign strategy. This occurs via the use of cross tabulation. Crosstabs breaks down polls results by variables such as gender, demography, age, location OR any other variable that the campaign deems to be important.
Years ago, I found that ethnicity correlates poorly with voting behaviour in the culture capital Dangriga. But this may not be so in other parts of the country. Or in the USA where race and religion could be deterministic factors in local races. Crosstabs provides the data that campaigns can use to easily target their actual supporters, potential supporters and reachable voters. Resources are thereby used more efficiently and effectively.
In Belize, many voters are becoming less comfortable with publicly revealing their political preferences. This also occurs in the USA and is manifested by historically low voters turnouts. In the USA though, because apparently one can actually register to vote and cast a ballot on election day, it becomes possible to target marginal voters who generally didn’t engage in politics, or distrust institutions, or have low civic engagement. Some marginal voters or nonvoters may also be single issue voters and with targeting facilitated by crosstabulation, they can reached even on election day itself.
In the run up to the US elections, a telling statistic was the approval rating of President Trump. Basically, an approval rating poll provides a barometer of the national mood. His approval rating exhibited a continuous downward trend and at election day hovered in the low 30s. A similar approval rating poll of Prime Minister Briceno would also provide insights into the national mood.
Finally, exit polling. Not yet done in Belize. Exit polls in the USA showed that the single issue that was most dominant in the voters’ decision was Affordability/Cost of living.
Thank you for your attention to this matter.
Dr. Philip Castillo is a development economist and independent pollster. He can be reached at pjcastillobz@gmail.com