Campaign promises often capture headlines, but do they actually move votes? This column examines whether a promise to expand welfare for some mothers in Germany could swing short-term electoral support. The campaign pledge boosted the party’s vote share by over 12 percentage points among voters who stood to benefit. However, this boost vanished soon after the pledge was fulfilled. Voters may respond to credible promises of financial benefits, but their allegiance is temporary and does not translate into long-term party loyalty. It highlights the risk of parties prioritising short-lived electoral gains over durable policy commitments.
Politicians often wager their campaigns on big-ticket promises. Donald Trump rode to power in 2016 vowing to build a US–Mexico border wall, while Joe Biden’s 2020 campaign highlighted sweeping student debt relief. In the UK, the Brexit campaign famously promised to “take back control” and divert £350 million a week to the National Health Service (NHS) – a claim plastered on the side of a bus. Such economic pledges capture headlines and voter attention. But how much do they actually move votes, especially before any promise is delivered? Recent research sheds new light on this question.
In a new study (Ganslmeier 2025), I exploit a natural experiment in Germany to show that a well-timed welfare promise can indeed swing short-term support, boosting a party’s vote share by over 12 percentage points among benefiting voters. Crucially, however, this boost proved fleeting, vanishing soon after the pledge was fulfilled. The findings reveal a promise-fulfilment paradox: campaign promises can win votes upfront, but even under delivery, those gains vanish. Voters take the promise, collect the reward, and move on.
Promising pensions, reaping votes: A natural experiment in Germany
In the run-up to Germany’s 2013 federal election, the centre-right Christian Democrats made a striking campaign pledge known as the Mother’s Pension (Mütterrente). The promise offered extra pension credits to mothers whose children were born before 1992, recognising child-rearing years in pension benefits. This eligibility rule created a sharp dividing line: women with children born just before 1 January 1992 stood to gain financially, whereas those with children born just after that cutoff did not.
In effect, an accident of timing – a child’s birthdate – determined who would benefit, allowing researchers to isolate the promise’s causal impact on voting. I use a regression discontinuity design leveraging this cutoff, comparing otherwise similar mothers on either side of the birthdate threshold. This approach provides a rare chance to measure the pure effect of a campaign promise on voter alignment, holding constant other (un)observable factors.
Figure 1 Support for the conservative party around the eligibility cutoff
Notes: Mothers just eligible for the promised Mother’s Pension (children born before 1992) were substantially more likely to support the Christian Democrats than those just ineligible (children born after 1992). The discontinuity at the cutoff reflects the promise’s causal effect on party support (baseline regression discontinuity design estimate ~12–13 percentage points). Dots represent local averages; lines are fitted trends on each side of the cutoff.
The results are striking. Mothers who stood to benefit from the pension promise were about 12.7 percentage points more likely to support the Christian Democrats than demographically similar mothers who did not qualify. This is a large and politically meaningful effect: a swing of this magnitude can easily be the difference between winning and losing a closely fought election. Importantly, I find no mobilising effect on inactive voters – the pension pledge did not bring new voters to the polls, but it did cause already-engaged voters to switch their party preference.
Why was this pledge so persuasive? The Christian Democrats’ proposal was a targeted, tangible economic benefit, effectively putting money on the table for a specific group. The promise spoke directly to middle-aged women who had raised children decades earlier and felt left out of prior pension expansions. Many such women traditionally supported centre-left parties; offering them a pension boost gave the Christian Democrats a compelling reason to ‘trade loyalty’ for material gain.
The findings align with a broader literature on policy and voting: voters respond to credible promises of financial benefits, much as they do to actual policy handouts. Studies in other contexts have likewise shown that government transfers can translate into electoral support (Manacorda 2011, De La O 2013, Hidalgo and Nichter 2016). What my findings add is clear causal evidence that even a future benefit, merely promised during a campaign, can sway voters in the present – a form of prospective vote-buying.
A temporary realignment: Here today, gone tomorrow
While the pension pledge delivered an immediate boost at the ballot box, its effect proved short-lived. I track voters through subsequent election cycles and find that the bump in party alignment disappeared soon after the policy was delivered. By the time of the next federal election (2017), the voting patterns of eligible mothers had reverted to their pre-promise baseline. The Christian Democrats’ momentary gains among these voters evaporated once the Mother’s Pension was enacted and pocketed.
The promise was like a political coupon: valid once, quickly discarded after use. Voters took the deal on offer – a one-time vote in exchange for a benefit – but did not develop any lasting allegiance to the conservative party. This dynamic echoes the concept of ‘vote-share contracts’, where parties deliver targeted payoffs knowing full well the effect may expire once the benefit is redeemed (Gersbach 2007).
Figure 2 illustrates this transience. The black line plots the promise’s effect on alignment with the Christian Democrats over time, while the green line shows the effect on alignment with the Greens (the left-leaning party that lost the most support due to the pledge). In 2013–2015, the Christian Democrats enjoyed a large uptick in support among eligible mothers (about 12–15 percentage points).
Figure 2 Time-varying impact of the pension promise on party support
Notes: Each point shows the estimated treatment effect of being eligible for the Mother’s Pension on alignment with the pledge-making party (Christian Democrats [CDU/CSU], black) and a competitor party (the Greens, green) over different periods. During the 2013 campaign and shortly after, eligible mothers were 12–15 percentage points more likely to support the Christian Democrats (black line spikes upward), with a corresponding drop in Green party support (green line dips). The shaded area marks the campaign period. By 2017, the effect fades: the black line returns to ~0, and former Green supporters revert back to the Greens. Error bands denote 95% confidence intervals.
Conversely, alignment with the Greens among that group fell by a similar margin, indicating that many Greens voters switched to the Christian Democrats to obtain the promised benefit. However, the lines converge back to zero by the 2015–2017 period. The temporary converts returned to their prior political preference once the pension credits were delivered and the pledge was no longer salient.
Who was swayed? Economic insecurity and left-leaning voters
The allure of the pension promise was not uniform across the electorate; it worked best on certain subsets of voters. I find that the most economically insecure mothers exhibited the largest swings. Among eligible women of modest means – for example, those earning below €850/month (the ‘mini-job’ threshold), single mothers, or those with multiple children – support for the Christian Democrats jumped dramatically, by up to 25–30 percentage points. Figure 3 depicts these heterogeneous effects.
Figure 3 Who switched? Heterogeneous effects of the Mother’s Pension promise
Notes: Bars show the treatment effect on support for Christian Democrats for various subgroups of eligible mothers (with 95% confidence intervals). Economically vulnerable groups – those earning under €850 (low income), single/divorced/widowed mothers, and mothers with multiple children – saw very large voting swings (20–30 percentage points) in response to the pension pledge. More secure groups (e.g. married mothers or mothers with a single child) showed smaller, statistically weaker effects. Data from Ganslmeier (2025).
Why these differences? Economic insecurity makes voters more persuadable by welfare offers. For a mother scraping by on a small pension or low earnings, an extra benefit in old age represents a meaningful gain. My finding that poorer mothers were the most responsive is intuitive: for voters with tighter household budgets, a future pension boost – even years down the line – can offer tangible hope. This aligns with evidence that increasing household income can affect voting behaviour of low‑income groups (Simeonova et al. 2018).
Broader implications: Promises and the limits of persuasion
My study provides timely evidence of how transactional our politics can become. In an era of fractured loyalties and ‘floating’ voters, a well-crafted benefit promise is a powerful campaign tool – but also a fleeting one. Evidence from the 2016 US election suggests that such pledges can effectively channel the economic anxieties of the working class into temporary political support (Freund 2017). The German case is illustrative, and not unique.
Across Europe, parties of all stripes have increasingly turned to targeted social-spending pledges to win votes, from pension increases to tax cuts, especially as economic anxieties have grown. The lesson for political strategists is clear: campaign promises do work, but only while they’re fresh. A credible pledge of a material benefit can sway a segment of voters in the short run, even pulling them across ideological lines.
Yet the study also reveals the limits of promise-driven persuasion. Because the effect vanished after delivery, the pledge failed to reshape long-term party allegiance or ideology. In competitive democracies, this implies that parties can ‘rent’ votes but struggle to keep them once the transaction ends. The result may be a politics trapped in short cycles of give-and-take, with little incentive for long-term coherence.
There are accountability upsides. Voters do value follow-through: when policies match promises, they reward parties for keeping their word (Alesina et al. 2024). But if electoral payoffs fade quickly, the risk is that parties shy away from difficult but necessary commitments and instead escalate their pledges in pursuit of short-lived electoral gains.
At a systemic level, over-reliance on transactional politics can hollow out democratic trust. Populist leaders, in particular, often prioritise visible, short-term payoffs over durable policy (Woo-Mora 2024). As campaigns devolve into competitive auctions of targeted giveaways, voters may come to expect constant rewards, raising the cost of persuasion and fuelling disillusionment when delivery lags behind rhetoric.
The lesson is not to abandon promises but to reckon with their limits. Democracies depend not just on what parties offer, but on how credibly they govern once the applause fades. If every election is won with a bigger cheque, what happens when the account runs dry?
Source : VOXeu
