How Geography and Voter Turnout Shaped Zambia’s 2021 Presidential Election Outcome

 

The 2021 presidential elections in Zambia produced a highly differentiated electoral landscape characterised by regional political dominance, varying voter turnout levels, and the influence of the rural–urban voting dichotomy. Beyond the presidential results, constituency-level data reveals important trends regarding political party dominance, voter turnout, and a distinct electoral spatial footprint. An interrogation of 156 constituencies shows how the United Party for National Development (UPND) and the Patriotic Front (PF) performed across the country and how voter turnout patterns shaped the electoral landscape.

The constituency-level interrogation shows that UPND emerged as the dominant political party nationally (Table 1). Out of the 156 constituencies interrogated, UPND dominated 97 constituencies, representing 62.2% of the total. PF dominated 59 constituencies, accounting for 37.8%.

Table 1. Political Party Dominance by Constituency

Political Party	Constituencies Dominated	Percentage
UPND	97	62.2%
PF	59	37.8%
Total	156	100.0%

UPND’s electoral spatial footprint was more nationally distributed, with dominance across Southern, Western, North-Western, Central, Lusaka, and Copperbelt provinces, while PF demonstrated stronger dominance in Eastern, Luapula, Northern, and Muchinga provinces (Table 2).

Table 2. Constituency Political Party Dominance by Province

Province	UPND	PF
Central	62.5%	37.5%
Copperbelt	95.5%	4.5%
Eastern	16.7%	83.3%
Luapula	6.7%	93.3%
Lusaka	66.7%	33.3%
Muchinga	0.0%	100.0%
Northern	30.8%	69.2%
North-Western	100.0%	0.0%
Southern	100.0%	0.0%
Western	100.0%	0.0%
Total	62.2%	37.8%

The rural–urban dichotomy also played an important role in shaping the 2021 presidential election outcome. Rural landscapes constituted most of the constituencies, making them strategically significant in determining presidential election outcomes (Table 3).

Table 3. Rural–Urban Distribution of Constituencies

Area	No. of Constituencies	Percentage
Rural	114	73.1%
Urban	42	26.9%
Total	156	100.0%

Note: Rural-urban differentiation is based on the Zambia 2022 Census of Population and Housing (Population Summary Report Volume Two, Table 5.2. With the representative geographic location operational definition being where 50+1% of a household population resides.

UPND secured majority votes in 65 rural constituencies, representing 57.0% of all rural constituencies, while PF secured 49 rural constituencies, accounting for 43.0%. In urban constituencies, UPND secured 32 constituencies, representing 76.2%, while PF secured 10 constituencies, representing 23.8%. UPND’s stronger urban performance suggests that urban voter preferences demonstrated stronger support for UPND, while rural areas remained comparatively more competitive between the two parties.

Another notable pattern emerges from voter turnout and its association with party dominance. Voter turnout percentages were categorised into five classes ranging from the minimum observed turnout of 50.52% to the maximum of 89.70% (Table 4).

Table 4. Voter Turnout Classification

Class	Voter Turnout Range (%)
Very low turnout	50.52% – 58.35%
Low turnout	58.36% – 66.19%
Moderate turnout	66.20% – 74.03%
High turnout	74.04% – 81.86%
Very high turnout	81.87% – 89.70%

 

Constituencies that had very high turnout and high turnout were more frequently associated with UPND than PF. UPND dominated 51 constituencies within these turnout classes compared to PF’s 12 constituencies (Table 5).

Table 5. Voter Turnout by Political Party

Voter Turnout	UPND	PF	Total
Very high	20		20
High	31	12	43
Moderate	38	23	61
Low	6	22	28
Very low	2	2	4
Total	97	59	156

The concentration of UPND dominance within high-participation constituencies suggests that voter mobilisation constituted a major electoral advantage in the 2021 presidential election.

A further breakdown shows that most very high and high voter turnout constituencies were in rural landscapes (Table 6).

Table 6. Voter Turnout and Rural–Urban Dichotomy

Voter Turnout	Rural	Urban
Very high	100.0%	0.0%
High	74.4%	25.6%
Moderate	62.3%	37.7%
Low	71.4%	28.6%
Very low	100.0%	0.0%
Grand Total	73.1%	26.9%

 

Among UPND’s 51 constituencies within very high and high voter turnout categories, 41 were rural and 10 were urban. PF recorded 11 rural constituencies and one urban constituency within the same turnout categories (Table 7).

Table 7. Voter Turnout and Rural–Urban Dichotomy by Party

Geo-Location/ Turnout	UPND	PF	Total
Rural
Very high	20		20
High	21	11	32
Moderate	18	20	38
Low	4	16	20
Very low	2	2	4
Urban
High	10	1	11
Moderate	20	3	23
Low	2	6	8
Total	97	59	156

 

The voter turnout data therefore suggests that constituencies with stronger voter participation were more commonly associated with UPND party identity during the 2021 presidential election. Lower-turnout constituencies, on the other hand, appeared more evenly distributed between the two major political parties.

In addition, a comparative analysis of voter turnout between the 2016 and 2021 presidential elections reveals that turnout patterns were not random or isolated across constituencies. The analysis produced a Pearson correlation coefficient (r) of 0.85, indicating a very strong positive relationship between voter turnout levels in the two elections (Figure 1).

Figure 1. 2021, 2016 Presidential Elections Correlation

 

This suggests that constituencies which recorded comparatively high voter turnout in 2016 also tended to record comparatively high turnout in 2021. The findings indicate continuity in constituency-level voter participation patterns over time, suggesting that electoral mobilisation and voter engagement may have become structurally embedded within certain constituencies.

Further, a broader historical analysis of provincial voter turnout trends since the 1991 presidential elections reinforces the argument that voter participation patterns in Zambia exhibit strong spatial continuity over time. Provinces such as Southern, North-Western, Copperbelt, and, more recently, Western Province have consistently recorded comparatively high voter turnout levels across multiple election cycles, culminating in exceptionally high turnout levels in 2021 (Figure 2).

Figure 2. Provincial Voter Turnout Trends (1991-2021)

 

The persistence of relatively higher voter turnout within these provinces suggests that voter participation may not merely reflect election-specific dynamics, but also historically embedded patterns of political mobilisation and electoral engagement. The findings therefore indicate that certain electoral spaces in Zambia have evolved into persistently high-mobilisation regions capable of exerting disproportionate influence on presidential election outcomes across successive electoral cycles.

Overall, the 2021 presidential election outcome was fundamentally shaped by the interaction between geography, voter turnout, and voter mobilisation. UPND’s electoral success was driven by its broader electoral spatial footprint across provinces, stronger performance in both rural and urban constituencies, and its clear association with high voter turnout constituencies. PF, while remaining competitive in several constituencies, demonstrated a narrower electoral spatial footprint, most evident from the election outcomes in Eastern, Luapula, Northern and Muchinga provinces.

The 2021 presidential elections therefore demonstrate how geographic reach and sustained voter mobilisation can become decisive determinants of presidential election outcomes.

In retrospect, several lessons emerge from the foregoing regarding political party competitiveness.

First, future presidential election competitiveness is likely to depend increasingly on broader electoral spatial footprints rather than geographically concentrated regional support.

Second, rural constituencies are likely to remain strategically decisive because they account for more than 70% of all constituencies nationally. Any political party seeking electoral competitiveness must therefore maintain strong rural penetration while simultaneously remaining competitive in urban constituencies, where voter concentration and turnout can significantly influence electoral outcomes.

Third, urban constituencies are likely to become increasingly influential in shaping future presidential election outcomes. UPND’s overwhelming dominance in urban constituencies suggests that urban voter consolidation can provide a significant electoral advantage. As urbanisation expands, political parties that fail to remain competitive in urban areas may experience declining national electoral competitiveness.

Fourth, historically high voter turnout constituencies are likely to remain strategically important battlegrounds in future elections. The strong positive correlation between voter turnout in the 2016 and 2021 presidential elections and the observed historical provincial voter turnout trends, suggests that spatial patterns of voter participation exhibit continuity across election cycles. Constituencies with historically high turnout may therefore continue to exert disproportionate influence on future presidential election outcomes, making them critical targets for sustained political mobilisation.

Fifth, regionally concentrated political support may become increasingly insufficient for securing national electoral dominance. The 2021 presidential election demonstrated that a broader electoral spatial footprints provided a stronger competitive advantage than concentrated regional dominance. Future presidential elections are therefore likely to reward political parties capable of expanding support beyond traditional strongholds and building broader national electoral coalitions.