Northern Ireland’s cycling puzzle: better satisfaction, little access and a car-dominated reality

Northern Ireland's latest walking and cycling figures contain a puzzle that is easy to misread and too important to ignore. In the 2025/26 Continuous Household Survey, 64 per cent of respondents said they would be likely to walk a short journey of up to three kilometres, while just 18 per cent said they would be likely to cycle a journey of up to five kilometres. The first impulse is to describe a 46-point gulf in public enthusiasm. But the two questions are not identical, and neither is a count of journeys actually made. The more interesting story lies in what the survey measures, what it leaves unresolved and how its results compare with Northern Ireland's record of everyday travel.

Published by the Department for Infrastructure on 5 November 2026, the statistics give policymakers a fresh set of signals about the conditions in which people might choose active travel. They do not provide a simple referendum on cycling, nor do they prove that a particular infrastructure scheme would convert a specified proportion of drivers. Read carefully alongside the 2024/25 survey and the separate Travel Survey for Northern Ireland, however, the figures reveal a transport challenge that deserves something more than the usual press-release arithmetic.

An apparent gap, with an important qualification

The latest household survey asked about willingness to walk journeys up to 3km and to cycle journeys up to 5km. Its headline responses were 64 per cent and 18 per cent respectively. The gap is 46 percentage points, but the difference in journey length, the activities involved and the exact wording means it is not a controlled comparison of the same trip. Walking two miles and cycling three miles are different choices, made under different conditions by people with different abilities and equipment.

There is another reason to resist an exaggerated conclusion. Asked whether particular activities counted as active travel, 88 per cent of respondents recognised walking and 67 per cent recognised cycling. That difference may relate partly to understanding of the term 'active travel', but the survey headline does not tell us why respondents answered that way. It would be an unsupported leap to claim that people who did not select cycling are opposed to it or do not know how to ride. The value of the statistic is that it shows even the language of transport policy is not understood identically by everybody.

The survey also reported 62 per cent satisfaction with local walking conditions, compared with 51 per cent satisfaction with local cycling conditions. Again, these are attitudes towards the respondent's local area, not technical audits of pavements and cycleways. Somebody can be satisfied with a particular path yet rarely use it; somebody can ride frequently but judge their local facilities poor. Satisfaction should not be substituted for safety, usage or network continuity.

One year's movement is not a revolution

The department's previous 2024/25 release offers a useful comparison. It reported 64 per cent likely to walk short journeys—the same headline level as 2025/26—and 16 per cent likely to cycle short journeys, versus 18 per cent in the new publication. Local cycling satisfaction was 48 per cent, versus 51 per cent now. Walking satisfaction was 61 per cent, compared with 62 per cent in 2025/26.

On the face of it, that is a two-percentage-point increase in willingness to cycle and a three-point increase in satisfaction with cycling conditions, alongside stable walking willingness and a one-point rise in walking satisfaction. But those simple differences are not proof of a statistically significant improvement. The published press summaries do not establish here the confidence intervals, effective sample sizes for each question or whether methodological differences affect comparisons. To pronounce a national trend from those figures alone would be more confident than the available evidence permits.

The earlier 2024/25 survey was based on 3,790 households and 4,584 individual respondents, according to DfI. The new survey drew from 3,459 households and 4,004 individuals. Yet the 2025/26 department release explicitly says the walking and cycling questions were asked only in the even-numbered questionnaire versions of a split sample, so those headline totals must not be presented as the number of responses to each cycling question. The survey's precise bases and weights belong in any rigorous interpretation of apparent year-on-year change.

DfI also states that the 2025/26 interviews were primarily face to face, with telephone interviewing available. It warns readers to take particular care comparing years in which pandemic conditions altered the interview approach. A chart spanning multiple years can look like a smooth behavioural trend while hiding changes in fieldwork. That is why statistical method is not an afterthought to good transport reporting; it is part of the story.

The bicycle-access denominator matters

One especially tempting figure is that only 31 per cent of 2025/26 survey respondents reported access to a bicycle. Among those who owned or had access to one, 30 per cent had cycled in the previous four weeks. Those measures describe different groups. The second is not '30 per cent of Northern Ireland adults cycled recently', because it is conditional on access. Likewise, the 31 per cent access figure is not the share of households with a bicycle unless the report defines it that way; the release refers to respondents.

A crude multiplication of 31 per cent by 30 per cent produces 9.3 per cent, which might look like a neat national participation estimate. It should not be published as an official finding. Survey weighting, routing and base definitions need checking before such a calculation is treated as representative. The proper finding is narrower and informative: bicycle access is limited, and recent cycling is not universal even among people with access. An infrastructure discussion that ignores who has a usable bicycle—and who cannot ride—will miss part of the problem.

For a policy reader, this suggests distinct groups requiring different responses. Some people already cycle and mainly want safer connections. Some may ride occasionally but not use a bicycle for practical journeys. Others lack access to suitable equipment, secure storage or the physical ability to cycle. Those are hypotheses to investigate, not individually measured explanations in the press summary. Different barriers demand different interventions; a single national percentage cannot identify their relative importance.

Willingness is not behaviour: a second survey changes the picture

The Travel Survey for Northern Ireland gives another view, drawing on journeys actually recorded rather than hypothetical willingness. Its 2024 figures, released in August 2026, estimate an average of 857 journeys per person during the year, of which 591 were by car. Car trips accounted for 69 per cent of journeys. Walking and cycling together accounted for 204 journeys per person, or 24 per cent, while public transport accounted for 37 journeys, or four per cent.

These categories do not form a simple translation of the Continuous Household Survey's willingness percentages. The travel survey covers actual journey behaviour; the household survey asks about attitudes and likelihood for specified distances. Their sampling methods, periods and definitions need to be respected. But together they show the question policymakers should confront: whether a willingness expressed in principle can become a practical journey when people have destinations, time constraints, children, equipment and traffic to contend with.

Another data point illustrates the challenge more sharply. DfI's March 2026 report on pupil travel says 64 per cent of primary school pupils were usually driven to or from school, 27 per cent walked or cycled and nine per cent took a bus. Among primary pupils living a mile or less from school, 52 per cent were driven and 48 per cent walked or cycled. Those figures are not evidence that all short car school journeys are unnecessary. Family schedules, road crossings, disabilities and other responsibilities matter. They do establish that proximity alone does not dictate mode choice.

The school evidence also warns against treating every trip as a free-standing decision. A parent might drive a short school journey because it fits a longer onward commute; another may walk because a safe crossing exists. The same physical distance can produce different choices. A serious NI transport investigation should examine trip chains and local conditions rather than assume every journey under a certain length is automatically replaceable.

What the figures do not prove about safety

There is a common argument that more protected cycling infrastructure would produce more riders, and another that facilities remain underused because people prefer their cars. Neither can be conclusively tested using these headline attitudes alone. A 51 per cent local satisfaction rating does not count collisions or near misses. An 18 per cent willingness figure does not tell us how many people would change their behaviour after a new crossing opens. The more rigorous test requires route-specific before-and-after counts, surveys that ask about barriers, and data on who uses a facility at different times.

Sailortown's proposed 0.8km link between York Street Station and the Cathedral Quarter is an example of the question. If it addresses a previously difficult junction and joins to other safe sections, its value may extend beyond its length. If it terminates abruptly, it may serve existing confident cyclists more than new riders. The survey provides context for why network quality matters; it cannot predict how many people will use the proposed route. The same applies to the 1.7km Newcastle–Castlewellan scheme now planned for construction. A national attitude figure is not a forecast for County Down.

DfI's Belfast Cycling Network Delivery Plan, published in 2022, envisages more than 50 schemes and emphasises connecting isolated infrastructure. That aim is consistent with the idea that a journey is only as useful as its weakest connection, but the official survey does not test the plan's success. To assess delivery, readers would need information on completed links, outstanding gaps, usage patterns and accessibility. Without those, apparent improvement in survey sentiment should be described cautiously.

Three questions for policymakers

First, what is holding back the willing-but-inactive group? The household release tells us the share who say they would be likely to cycle a short distance, and whether respondents report recent cycling if they have bicycle access. It does not reveal a single dominant cause for non-participation. More detailed tables may establish differences by demographic characteristics, but those should be checked, not guessed from stereotypes about rural residents or age groups.

Second, what counts as progress? An increase in satisfaction might be encouraging even if overall cycling remains uncommon, while a new route might improve safety without creating a large immediate modal shift. DfI should specify whether it is trying to increase journey share, improve access, reduce collisions, support leisure activity or make the network usable for a wider population. Those goals overlap, but they are not identical and they require different measures.

Third, how will trade-offs be assessed? Road space is finite in established towns and cities. Changes to crossings, loading, parking and traffic signals can have costs and benefits for different users. Those impacts should be measured against clear objectives rather than argued from national headline percentages. A policy claiming to be evidence-led needs to explain which evidence supports a particular design and what would count as success or failure after opening.

The finding behind the figures

PureCar's interpretation is that Northern Ireland has evidence of a substantial distinction between hypothetical willingness to travel actively and the routines recorded in actual journey surveys, but it does not yet have an explanation of that distinction from these headline figures alone. That is more defensible than describing residents as simply unwilling to cycle. The latest survey is useful precisely because it invites better questions about access, safety, continuity and the competing purposes of a journey.

There are signs of modest improvement on some reported cycling-attitude measures between 2024/25 and 2025/26. Their statistical significance needs checking before turning them into a trend. Bicycle access remains a separate consideration from recent use. Journey behaviour remains dominated by cars when measured across Northern Ireland, even while many people say they could contemplate short active journeys. None of those findings is contradictory once the surveys are read on their own terms.

The most useful test of Northern Ireland's investment in walking and cycling will not be whether next year's press statement contains a higher percentage. It will be whether officials can show that specific barriers have been removed, people with different needs can use the facilities, and actual journeys have changed where the evidence justifies expecting change. Numbers deserve more than being decorated with policy slogans. They deserve to be interpreted accurately enough that the next decision can be judged against them.