By Zambian Economist Analyst
Campaign season produces a lot of crowd photos and very few reliable conclusions. With rallies drawing large turnouts across the country ahead of the 13 August election, it is worth asking, as a matter of economics and behaviour rather than politics, whether crowd size actually tells us anything about how people will vote. Zambia’s own recent history suggests the answer is: not much on its own.
The clearest evidence is already in the record
In the 2021 general election, then-President Edgar Lungu drew large crowds throughout his campaign. He lost decisively to Hakainde Hichilema, who won with just over 59% of the vote against Lungu’s roughly 35.6%, on a turnout of 70.61% of registered voters. That result is one of the clearest domestic illustrations available that rally attendance and vote share can move in entirely different directions. It does not prove crowds are meaningless, only that they are not, by themselves, evidence of anything in particular.
Why crowd size is a weak predictor: four economic reasons
- Selection bias. The people who turn up to a rally are disproportionately existing, motivated supporters, not the undecided voters who actually determine close contests. A large crowd mostly demonstrates that a campaign can mobilise its base, which is necessary but not sufficient for winning.
- The handout economy. Attendance is not a pure expression of political preference when it is subsidised. Branded T-shirts, chitenges, caps and sometimes transport or food are distributed at rallies, the same campaign spending we examined in our piece on the economics of campaign merchandise. Some share of any crowd is there partly for the merchandise or the outing, which inflates headcount without reflecting commitment at the ballot box.
- Curiosity and spectacle. Public figures draw onlookers regardless of political alignment, particularly in smaller towns where a high-profile visit is a rare event in itself.
- Geographic and logistics bias. Rallies happen where roads, venues and security arrangements make them feasible, not in proportion to the electoral map. A campaign can look dominant in the towns it visits while doing far less visible work in the constituencies that actually decide the national count.
What the arithmetic actually requires
Zambia’s presidential election is decided by a simple constitutional threshold: more than 50% of the valid vote. The Electoral Commission of Zambia has certified 8,786,300 registered voters for the 13 August 2026 poll, spread across provinces led by Lusaka (1,430,889) and Copperbelt (1,296,446). In 2021, turnout was 70.61%, or 4,983,099 votes cast out of registered voters at the time. No rally, however large, changes that underlying arithmetic; it can only be read, at best, as one noisy signal of organisational capacity in the specific place it happened.
Some commentary this cycle has floated turnout estimates for 2026 well above the 2021 figure, built partly on qualitative observation of rally attendance rather than polling. That kind of forecasting is worth treating with real caution: the same commentary acknowledges that crowd size is, in its own words, an imperfect proxy for voter engagement, given multiple attendances by the same individuals, organised mobilisation, and the incentives described above. Zambian Economist does not publish or endorse vote or turnout projections of that kind, official or otherwise, and neither should a reader treat them as more reliable than they are.
What actually does correlate with results
Where rally coverage runs out, ground organisation matters more: structures at polling-station level, across all 13,529 stations nationally, door-to-door mobilisation, and the unglamorous logistics of getting registered supporters to actually cast a ballot on the day. That is consistent with the standard finding in political economy research more broadly, visible, high-attendance events build momentum and media coverage, but the deciding factor in close elections is almost always organisational reach, not rally size.
The economics of running a rally
A large rally is itself a campaign expense, not a free event. Sound systems, stage hire, security, transport for supporters bussed in from outlying areas, and the printed materials distributed on site all draw from the same finite campaign budget covered in our piece on how Zambian businesses should plan for the election period. A campaign that spends heavily to fill a stadium in one town has, by definition, less to spend on the polling-station-level organisation that research suggests matters more on election day itself.
Common mistakes when reading rally coverage
- Treating a crowd photo as data. Attendance estimates from photos and video are notoriously unreliable and easily disputed by rival campaigns.
- Extrapolating nationally from a handful of high-profile events. A strong showing in a few large towns says little about turnout in the hundreds of rural constituencies that receive no rally at all.
- Ignoring the subsidy effect. Free merchandise, transport or food changes who shows up and why, without changing how they vote.
- Confusing enthusiasm with certainty. Zambia’s own 2021 result is the clearest reminder available that the two are not the same thing.
Actual turnout on 13 August against the 2021 baseline of 70.61%, and how it varies by province against the registered-voter numbers above. See our 2026 economic calendar for key dates through polling day.
Do large campaign rallies predict election results in Zambia?
Not reliably. In 2021, Edgar Lungu drew large crowds throughout his campaign but lost decisively to Hakainde Hichilema, illustrating that attendance and vote share can diverge sharply.
Why do people attend rallies even if they are undecided or unsupportive?
Reasons include curiosity about a public figure, subsidised transport or food, and free branded merchandise distributed at events, none of which necessarily reflect a firm voting intention.
How many registered voters does Zambia have for the 2026 election?
8,786,300, according to the Electoral Commission of Zambia, up from the 2021 registered voter base.
What turnout was recorded in the 2021 election?
70.61% of registered voters, or 4,983,099 votes cast.
What vote share does a candidate need to win outright?
More than 50% of the valid vote, under Zambia’s two-round presidential election system.
Are unofficial election or turnout projections reliable?
Projections based on qualitative observation of rally crowds rather than scientific polling should be treated with caution. Zambian Economist does not publish or endorse such projections.
What matters more than rally size for winning an election?
Ground-level organisation, structures at each of Zambia’s 13,529 polling stations and door-to-door mobilisation, is generally considered more decisive than the size of any single rally.
