VAP vs VEP Voter Turnout Calculation & the Dark Side of Statistical Figures
Democracy, by definition, is a popular concept and, in the theater of politics, a popular concept is an important one. States usually take care to maintain democratic appearances with their populations and the international perception of democracy is often what determines whether the press refers to a nation’s leaders as a government, a regime, or a dictatorship. Since it is impossible to ask a whole population whether it feels enfranchised or not, the integrity of democracy must be inferred with statistics by looking at voter participation, exit polls, public-surveys, and other data. Of these, one of the more important is voter-turnout and — with mass media being such a minefield of dishonest statistical figures and the 2018 midterms being just around the corner — it can be handy to know how turnout is calculated…
How to Calculate Voter-Turnout & Fight the Dark Side of Statistics
Without voter-turnout, it would be hard to judge whether an election is meaningfully democratic because turnout is what tells us how much of a population actually participated in the collective decision-making process otherwise known as democracy. If just 10% of a country’s people votes to elect their leaders, for instance, only a cynic would call it democracy — but if 90% participates in the elections, it would be hard to characterize it as anything else. Turnout isn’t everything, of course — in a situation where a state only allowed one candidate to run, for example, or where systematic violence was used to bully people into voting a particular way, such elections could not be democratic no matter how many ‘voted.’
But so long as no disqualifying anti-democratic shenanigans occur, turnout is pretty useful for evaluating the quality of elections and there are two basic ways to measure it — one uses the VAP or voting-age population and the other uses the VEP or voting-eligible population.
Voting-Age vs Voting-Eligible Population: a Critical Comparison
Calculating voter-turnout using a nation’s VAP is pretty straightforward — the number who voted is simply divided by the number who are old enough to vote. It is a simple fraction, like so:
Number of Votes / VAP = % Turnout
When turnout is calculated using the voting-eligible population or VEP, the number of people who voted is divided by the result of subtracting the people designated by the state’s authorities as non-eligible to vote from the total of the voting-age population. Some algebra, along with some research into the state’s electoral laws, is needed to figure out the VEP-based turnout:
Number of Votes / (VAP — Non-Eligible Population) = % Turnout
Without getting too entangled in the gazillions of non-essential details, these are the two major approaches to turnout — only one of them is useful, however.
How Measuring Voter-Turnout w/ VEP Is Misleading
To illustrate why VEP should never be used to calculate turnout, consider the fact that Saudi Arabia kept women from voting until just 3 years ago in 2015. Turnout figures calculated with VAP would clearly show that half the people were disenfranchised by the Saudi monarchs — a VEP-based turnout figure, on the other hand, would suggest Saudi Arabia was a functional democracy. And in the technical terminology used in the field of statistics, this type of thing is known as horseshit. Using VEP to measure turnout basically lets a nation’s political-authorities customize the equation that is meant to determine whether its elections are “free and fair” — a flaw so spectacularly fatal that it defeats the whole purpose of turnout before the measuring even begins.
Case in Point: VEP vs VAP Turnout, 1989 South African Elections
As a more concrete example, consider the turnout rate for the 1989 elections held in apartheid-era South Africa. At the time, only whites were eligible to vote and, with this nation being located in Africa and everything, this policy tended to exclude quite a lot of people, by which the author means almost everyone. If the South African voting-eligible population is used, then it appears that there was an impressive turnout for 1989 because a total of 2.2 million or about 87% of 2.5 million suspiciously-pale voting-eligible South Africans turned out to vote. See:
2,157,593 / 2,478,930 = 0.87037 or about 87%
In reality, of course, the 1989 elections were a reprehensibly illegitimate farce — the actual turnout rate, calculated honestly by using South Africa’s voting-age population¹ as the denominator, was really just 12.7% —
2,157,593 / 17,047,560 = 0.1265 or about 12.7%
The Distortion of US Voter-Turnout Today
Though apartheid South Africa and the Saudi regime in Arabia are some of recent history’s more dramatic cases, the erasure of whomever election-authorities label ‘ineligible’ from voter-turnout figures is so widely accepted among the mass media commentariat that the legitimacy and usefulness of such calculations is rarely questioned. This enables countries like the US to erase a few percentage-points from its population — mostly People of Color — with the flick of a pen by denying voting-rights to incarcerated citizens, along with a few million more felons disenfranchised by state-level laws (most of whom are also People of Color).
6.1m disenfranchised felons & prisoners / 245.5m VAP = 0.0248
or about 2.5%
And 2.5% of the US-American adult population is erased — just like that. 7.7% of Black US adults, erased — just like that. The disenfranchisement of these unlucky, mostly non-violent, mostly Black and Brown people is not accounted for in the VEP-based figures that US news-media seems to favor. Among many others excluded from these statistics are about 22 million lawful residents and other human-beings whose citizenship-related fees and paperworks have yet to find their way through the bureaucratic labyrinth of the US immigration process.
22m / 245.5m = 0.0896 or about 9%
And another 9% of the US adult population — along with the 2.5% erased by incarceration or past conviction — disappears into the statistical void of ineligibility and, at the same time, US voter-turnout grows by as much as 11.4%.
VAP Is Ethically & Empirically Superior to VEP
In addition to providing a way for despotic states to edit their own turnout statistics by restricting eligibility to whomever they please, the use of VEP also sabotages the scientific usefulness of turnout data. Since demographic traits like age exist in populations of all nationalities, VAP-based turnout in different countries can be compared, allowing scientists to study how different forms of democracy and election-systems may affect levels of voter-participation. Since ‘eligibility’ is a legal construction that differs depending on the political-leadership of various nation-states at any historical moment, turnout measured using VEP cannot be compared with data from other countries or even to the same country’s past and future turnout because eligibility requirements (like most laws) tend to change over time.
Real Voter Turnout Is Always VAP, VEP Only Distorts It
At the time of writing, the 2018 US midterm elections are only weeks away and that means we are all about to hear a ton of analysis and hot takes from the news-media commentariat, much of which will inevitably deal with the impact of voter turnout. Many of these numbers will ignore the tens of millions of disenfranchised people living in the land of the free because ignoring them is what allows everyone else to continue believing they live in some kind of democracy or, at the very least, an acceptably-democratic republic. But this is a fantasy that cannot withstand an honest look at the numbers. For the sake of those whose voices are silenced to prop up the fantasy, it would be good if those whose voices are still audible would use them to call out the collaborators whose bad statistical analysis helps to make such a farcical democracy possible.
In solidarity,
John Laurits
Notes:
¹ Due to a lack of reliable demographic data for age distribution in the South African population prior to the 1994 elections, I’ve estimated the South African VAP for 1989 based on the ratio of voting-age South Africans to the total South African population 3 years later in 1994. My figure for the 1989 VAP is based on what I judge to be the reasonable assumption that the VAP-to-total-population ratio did not change significantly in the 5yr period from 1989–1994. For transparency’s sake:
(1994 est. pop / est. VAP) 23,063,910 / 40,436,000 = 0.57038, or ~ 4:7 and ~ 57% ;
(1989 est. white pop. X 0.57) 4,349,000 X (0.57) = eligible white population of 2,478,930 or about 2.5m
