FixMyStreet!

The geography of citizen reporting on neighbourhood issues in the UK

Other formats: Plain text, PDF, EPUB.

This short report accompanies a set of online maps and data produced as part of a collaboration between researchers at the University of Stirling, the University of Sheffield and mySociety (a not-for-profit social enterprise). The work was funded by the Consumer Data Research Centre (CDRC), which was established by the UK Economic and Social Research Council as part of phase two of the Big Data Network.

In this report we provide commentary on our attempt to understand spatial patterns of citizen reporting on local issues though mySociety's FixMyStreet app and website over an eleven year period. At the outset, it is important to be aware of a number of methodological caveats that must be borne in mind when interpreting the results. These are mentioned throughout the report, and in the penultimate section in particular in a series of Frequently Asked Questions.

Put simply, we would urge readers not to jump to conclusions about local environmental conditions based on our results alone. There are many reasons for variations in the reporting of neighbourhood issues, such as differing levels of awareness about FixMyStreet between areas, and the propensity of different groups of people to report problems. It is well known that some people report more than others, so this must be taken into account when interpreting the data.

What we can say for sure from our results is where the reports are, how many there are, and what category they fall into. Therefore, we suggest our results are most useful for comparing reporting between areas in relation to i) volume and ii) report type. Nonetheless, we also believe there is significant value in looking at the micro-geographies of neighbourhood fault reporting across the UK because it often provides an insight into what local residents find important in relation to their local environment. It can also provide important information on the differences in reporting rates between areas.

What we found is that over an eleven year period from 2007 to 2017 more than 1.1 million reports about neighbourhood environmental problems were submitted to local authorities via FixMyStreet. These reports are not uniformly distributed across the UK, nor are they uniformly distributed within local authorities. We see clusters of reports in some neighbourhoods more than others and, in relation to deprivation, there are proportionately fewer reports in ‘decile 1' areas (the most deprived areas) and most in ‘decile 7' (see below). Most reports relate to things like potholes (about half of all reports nationwide) and environmental health issues like rubbish on the streets and dog fouling (about a quarter of all reports).

In order to make sense of the many types of issue that are reported via FixMyStreet, the research team developed a classification system which divided reports by broad type, as follows: ‘Road Safety & Defects', ‘Environmental Health', ‘Abandoned Vehicles & Parking', ‘Environmental Disruptions', ‘Public Spaces', ‘Incivilities', ‘Access' and ‘Other'. You can read more about these categories on the project website.

Introduction

Providing local environmental services - e.g. repairing streets and collecting rubbish - is one of the most basic tasks of local government. Their importance to citizens can be seen through the regular stories of "angry people in local newspapers" complaining about poor services.

Yet because of their mundane nature they are often not taken seriously as a policy issue, with greater focus on "serious" issues such as child protection or social care. As concerned citizens and their representatives will highlight, though, these services and good local environments matter a great deal for people's sense of pride in their community.

In fact, wider research also highlights the links between a good quality local environment and good health and wellbeing of residents, and also the social capital and empowerment felt by residents.

Why such services are of interest here is that they are commonly delivered through a response to a citizen-initiated request. This has become more common in recent years as budget cuts to local councils mean they have fewer resources to do regular inspections. Evidence also suggests that different citizens are more likely to make such requests - namely, that it is better educated, higher income residents who will contact their local council to have a problem fixed.

Therefore there is a risk that relying on citizens to report problems in their neighbourhood could lead to more affluent neighbourhoods getting a higher level of service from their local council. One way to assess this is to look at where requests for such services come from. This is part of what we attempted to do in this project.

For this project we analysed one partial set of data on such requests - the database of reports to local councils using the FixMyStreet website or app. We say ‘partial' because these reports do not, of course, include all reports received by councils across the UK, though in some areas they account for a significant proportion. Since 2007, over one million environmental problems have been reported to local councils across the UK through FixMyStreet. We do not know what proportion of total reports this accounts for.

By mapping this data onto neighbourhoods ranked according to deprivation indices used across the UK we have found that:

  • There were more problems reported in neighbourhoods in decile seven of the indices of multiple deprivation than other neighbourhoods;
  • There were more reports of problems such as littering and dog fouling in the most deprived neighbourhoods;
  • There were more reports of road defects in the least deprived neighbourhoods;
  • Reporting rates vary between local councils, with very limited use of FixMyStreet in Northern Ireland;
  • Across the UK, just over 52% of all reports were about road defects or road safety (typically potholes);
  • Environmental health issues (such as fly-tipping and dog fouling) accounted for just over 24% of all UK reports; and
  • Overall, there are clear differences between areas in relation to the kinds of things that are reported most frequently.

N.B. The most deprived areas are to the left of the X axis (e.g. D1, D2) and the least deprived areas are to the right (e.g. D9, D10).

What is FixMyStreet?

FixMyStreet.com is a website (see below) where anyone can report to their local council a fault, defect or problem with their local area. In addition to the website, there is also an app which people can use to report problems via their mobile phone. This can be done anonymously, if desired.

To give a more concrete example, let's consider the following scenario.

A resident of 29 Acacia Road (let's call them Eric) leaves their house one day and notices several bags of rubbish on the pavement, yet it is not bin day. One of the bags is broken and a dangerous banana peel has spilled onto the pavement. Bin day comes and goes yet the rubbish is still there, so Eric decides to use the FixMyStreet app on his phone to take a photo of the rubbish, which he now assumes has been fly-tipped, and then submits this via FixMyStreet, who then pass Eric's report on to the council, who then decide how to respond. A month later, Eric is sent a questionnaire to find out if the problem has been fixed (some councils have a direct integration into their own reporting system that updates the status of the report). Some people answer the survey, some people don't.

In summary, then, it works like this:

  1. User enters a nearby UK postcode, or street name and area
  2. User locates the problem on a map of the area
  3. User enters details of the problem (including a photo if they want)
  4. FixMyStreet send it to the council on user's behalf

We were provided with anonymised data for eleven full calendar years from 2007 to 2017 and, using the most recent FixMyStreet summary data, we can see that most reports do not show up as having been ‘fixed' on the FixMyStreet website (see below).

There are many reasons for this, of course. The problem may have been fixed, but the reporter didn't respond to the survey. It may be the case that what has been reported is not the council's responsibility (e.g. it may be the responsibility of a private landowner), or it may be the case that individual councils have to make very difficult choices about what to respond to and what to ignore, owing to budget cuts. These are just some of the more obvious reasons.

For the purposes of our research, focused as it is on the spatial dimensions of citizen reporting, the fact that all FixMyStreet reports have an accurate location associated with them means that we are able to locate them and map them in a systematic manner. Users of the website or app can also see current reports on a web map, as shown below for part of Edinburgh.

It is important to note that there are no limits on how many reports people can submit, and in some areas it would appear that a few prolific reporters are responsible for a majority of reports. However, this is very much the exception.

Finally, we need to point out that a small number of UK local authorities use FixMyStreet as their default reporting tool and so far show more reports. In our online maps these are referred to as ‘Co-Brands'.

What do we know about citizen reporting?

We know that not everyone regularly reports local issues to their local councils and we also know that certain characteristics of citizens shape these broader patterns.

For a report to occur we need a citizen to:

  • Notice and recognise a problem;
  • Know who to report it to, and how to do this (in our case, by using the FixMyStreet website or app);
  • Have the language skills to report it - a poorly worded, or difficult to understand report might not get a response;
  • Have an incentive to report an issue - it must either directly affect the citizen, or they must recognise that it will have a negative impact on other citizens.

A key variable we know that affects these factors (particularly the incentive to report) is the socio-economic status of a citizen - that is, people with higher education levels, higher status jobs, and/or higher household income are more likely to report issues.

Unfortunately, in the UK, we do not have regular survey data that records these factors among the population and also accurately records how often people report local issues. However, a recent synthesis of research across a range of policy areas, looking at different ways of engaging, in different national contexts, did suggest that the higher someone's socio-economic status, the greater the likelihood that they will report issues to their local council. We also know from research in the USA that local neighbourhood conditions also affect the likelihood that someone will report an issue.

Firstly, citizens are much more likely to report issues that are close to where they live and, possibly surprisingly, they are more likely to report these issues than issues where they spend significant amounts of their time (e.g. outside their workplace, or on regular journey routes).

Secondly, we know that the "broken windows" thesis applies in the case of the sorts of issues that FixMyStreet deals with. The original "broken windows" thesis suggested that if small problems in neighbourhoods are not fixed then people begin to think no-one cares, trust among citizens reduces, and this then leads to greater criminality.

The evidence of such an extensive link is not relevant in understanding this data, but there does seem to be a point where individual problems in a neighbourhood can accumulate to such a point that people think "nobody else cares, why should I?" and reporting rates drop-off. This may be part of the explanation as to why reports in areas which fall among the most deprived decile are significantly fewer than in all other categories.

On this topic, previous research by Hastings and Matthews has shown that more affluent residents tend to have advantages in public service provision of the kind we are investigating here; the so-called ‘sharp elbows' thesis. Further afield, in New York City, White and Trump have shown that similar data must be used with caution, and we echo those sentiments here. Yet we also agree that there is much to be gained from developing a deeper understanding of the data we explored, particularly in relation to its spatial manifestation and how areas with similar deprivation profiles compare with one another.

What do we know about local environmental issues?

We also know that neighbourhood conditions vary according to predictable patterns. Some of these are due to their nature: busy roads are more likely to have potholes reported (on many of the maps of the FixMyStreet data you can actually make out the road pattern of an area by these reports); areas with high footfall, like shopping streets, are likely to have higher reports of dropped litter and dog-fouling.

There might also be local conditions that are more difficult to understand systematically: the position of buildings, streets, and street furniture might create wind-traps that collect litter in particular spots; a neighbourhood might have a piece of land with unclear ownership which is regularly subject to fly-tipping. In short, local interpretation and local knowledge are very important when looking at this data at the neighbourhood level; less so at the level of the local authority.

Some systematic surveys of local environmental conditions have been carried out that provide us with some good evidence. Until 2015, the UK Government funded the Local Environmental Quality Survey in England (LEQSE) which sent surveyors to record a sample of streets across every local authority. The last results showed that problems such as littering, dog-fouling and fly-tipping were worse in the most deprived 10% of neighbourhoods (measured as census Local Super Output Areas). An equivalent survey in Scotland (the Local Environmental Audit Measurement System) found similar results.

We know some of the reasons for this. Population densities tend to be higher in the most deprived neighbourhoods, so there are more people to create litter. More lower-income households live in deprived neighbourhoods, and they are less likely to be able to afford durable goods or may have to buy secondhand goods which will need disposing of earlier. They are also less likely to be able to afford fees for disposal from local councils.

Due to the stigma associated with deprived neighbourhoods, other citizens may think that it is acceptable to fly-tip their waste in these areas. Finally, the design of some deprived neighbourhoods may make them more prone to environmental problems, for example: difficult to manage public open spaces, with unclear ownership; or building types that create wind traps.

From previous research we can suggest that deprived neighbourhoods are likely to have a greater concentration of problems, particularly litter, dog-fouling and fly-tipping, but citizens in these neighbourhoods are less likely to report such issues. This assumption is borne out in the data and can be seen in the maps we have produced for this project.

FixMyStreet outputs: an overview

Keeping in mind the caveats we have already mentioned, a UK-wide analysis of all eleven years of FixMyStreet data does reflect what we might expect from the research to-date:

  • Rates of reporting of problems like littering and dog fouling are higher in the most deprived neighbourhoods; less deprived neighbourhoods are much more likely to have reports of potholes.
  • The highest number of reports are in neighbourhoods in decile seven of nationwide deprivation indices (where one is the most deprived and ten is the least deprived).

This second finding might be explained by the nature of these neighbourhoods - these are typically suburban neighbourhoods with shared open space, such as play parks, so residents may be more likely to encounter problems to report in their everyday lives.

In order that our analysis was useful, and made sense at a local level, we decided to produce a high resolution map and data poster for every local authority in the United Kingdom, showing the location of reports, what category they fall in, how this compared to other areas and so on. We also compared the number of reports in each area to their deprivation profile, in order to get a sense of whether the level of reporting was higher or lower than might be expected.

There are 391 graphics in total - one for each UK local authority (except county councils as these overlap with other local authorities) - and we explain more about them here with reference to the data for Sunderland, in the north east of England. Each graphic covers the full period from 2007 to 2017 and relates to a single local authority area. The number of reports per 10,000 adults is also shown, in addition to a breakdown of reports by type.

We have also shown the percentage of reports received by year, so that interested parties can see better the uptake of FixMyStreet across the country. In the example of Sunderland below, it is only really from 2013 onwards that the number of reports increases significantly and then there is an obvious spike in reports in 2016. We can also see that almost 20% of all FixMyStreet reports in Sunderland are in decile 7, yet such areas account for only around 10% of all LSOAs in the local authority. The descriptive text on each poster provides more context and in this case we can see that Sunderland has a far higher proportion of Environmental Health reports than the UK average.

FixMyStreet: methodological FAQs

We want to be clear about what we think can and cannot be inferred from our analysis of FixMyStreet data, so we are providing answers to questions that were asked frequently during the course of the research project; either by the research team or by local authority consultees we shared the outputs with. The reason for doing this is that we want our results to be understood in context. We also want to make sure that the results are not misinterpreted and that we explain some potential pitfalls in the interpretation of results.

What do your maps tell me about the quality of my local environment?

Our maps provide details of where FixMyStreet reports are located over an eleven year period. The data released as part of this project can be broken down into individual years, but it is important to be aware that our data covers a relatively long time period so we encourage users to think of our outputs as a summary of neighbourhood fault reporting rather than an up-to-date statement on current neighbourhood conditions.

Why do your maps cover an eleven year period?

Across the UK there is considerable variation in the take-up of FixMyStreet so we decided to produce a set of maps that covered the whole period that FixMyStreet has been in operation, in order to provide an account of how local users have engaged with this service. In our graphics we have also included information on the percentage of reports by year so that users can see how this has changed over time. It is also important to note the reporting rate per 10,000 people and compare this to other local authority areas.

There are a lot of dots in my area: does this mean my neighbourhood is worse than others?

Maybe. Maybe not. What we can say about local areas with high levels of reporting is limited. We can be sure that the reports generated are real and that someone has been motivated enough to submit them but without further local analysis it is difficult to say for sure whether this is a reflection of a poor local environment or whether it is a reflection of a higher propensity to report on behalf of local residents. One way to ‘sense check' this, however, is to look at areas with similar deprivation profiles and see how they compare.

Doesn't the pattern of reporting say more about the underlying urban fabric than anything else?

Perhaps, and this is an important point. The underlying geography of houses, roads, parks and so on is an important factor in all of this. There will inevitably be a higher number of reports about, say, potholes, if there is more road surface in a particular area compared to another. If it is also the case that an area has a very high daytime population (e.g. it receives thousands of commuters and hence high footfall) then it is likely to see much higher levels of reporting than would be generated by residential population alone. This is why we urge users to investigate the data themselves, in addition to looking at our map and data outputs.

What about duplicate reports? Do you include them?

Yes, we include duplicate reports in the sense that we include multiple reports about the same issue; whether they are submitted by multiple individuals or the same individual more than once. There are two reasons for doing this. The first is that each report represents a form of citizen engagement, whether it is a new issue or a repeat report, and that is what we are trying to capture here. The second, is that it is the level of repeat reporting may provide a way to gauge the extent to which individual problems are perceived by local residents. That is, if an issue is reported multiple times by multiple users then we can reasonably infer that it is seen as a more serious issue than one that is not. Our analysis did not seek to explore this further but we think it is important not to exclude any data points at the outset.

What does a high level of reporting indicate?

There are a number of possible interpretations here. One is that a high level of reporting is indicative of digitally savvy local citizens who care about their neighbourhood and want to make a contribution to preserving its quality. Another is that a high level of reporting is indicative of a poor quality local environment. Our analysis of the data in addition to some local follow-up suggest that a mix of the two is more likely. In some areas, such as the most deprived parts of the UK, there are relatively low levels of reporting yet environmental quality is often visibly poor (e.g. this can be seen on Google Street View, or by walking around the neighbourhood). In such cases low levels of reporting should not be taken as an indication of a high quality neighbourhood. On the other hand, there are some areas (e.g. ‘decile 7', as discussed above) where reporting levels are relatively high yet the visual evidence suggests neighbourhood environmental quality is generally good. So, high levels of reporting may suggest higher expectations on the part of individuals more than anything else. Conversely, low levels of reporting may in some areas reflect low expectations about neighbourhood quality in the first place. These issues are things we think need to be investigated further.

Why are there no map outputs for county councils?

County councils overlap in area with district councils - while FixMyStreet sends different kinds of reports to the type of council responsible, for the purposes of these maps all reports are shown at the district level.

What do the maps and data actually tell us, then?

Our outputs provide new detail on the level, location and type of neighbourhood fault reporting across the UK. They allow us to make comparisons between local authority areas and between different kinds of areas within local authorities. They provide an initial insight into neighbourhood fault reporting as a starting point for further investigation.

Does a high reporting level of dropped banana peels suggest a superhero lives nearby?

We do not believe we can make such inferences about individual citizens in a neighbourhood from the dataset.

What does all this tell us and what have we learned?

We have published this short report as a guide to our outputs and as a cautionary note against jumping to conclusions about what gets reported and what does not. In terms of what our analysis of FixMyStreet data actually tells us and what we have learned from all this, we think the following points are worth emphasising.

  • We can see in detail where more than one million neighbourhood environmental quality reports have been submitted in the UK over more than a decade.
  • We can understand more about what local residents find important in relation to neighbourhood quality. By looking at the break-down of reports by type across different areas we get an insight into what matters to people. This predominantly seems to be things like potholes, fly-tipping and dog fouling.
  • We can begin to get a sense of how the uptake of the technology itself (i.e. the FixMyStreet website and app) has developed in different parts of the country.
  • We can begin to understand how reporting rates vary between different types of neighbourhood. In our study we used neighbourhood deprivation levels as a proxy for neighbourhood type, but of course there are many other ways to understand neighbourhoods. Nonetheless, we did observe some important differences in reporting levels in relation to deprivation, both at the national and local level.

Our study represents a first step towards a better understanding of the above issues. Although we urge readers to approach the data with caution, we think the results will be both interesting and useful for local people, councils and anyone with an interest in neighbourhood environmental quality. Interested parties can also download the data which mySociety have made available as part of this project, and then conduct further analyses.

Our hope is that through this initial piece of research it will be possible to understand better the ways in which citizens engage with neighbourhood fault reporting, how this varies by location and area type, and how this then feeds into decisions about what gets fixed. In relation to the latter, we hope this could be a useful way for councils to better understand whether service provision is equitable and neighbourhoods are assessed on the basis of need rather than where the residents are most ‘shouty' (e.g. ‘decile 7').

From what we know about neighbourhood environmental quality in the nation's most deprived areas, our analysis of FixMyStreet data would seem to suggest that there is significant under-reporting of issues in those neighbourhoods which fall into the most deprived decile.

Acknowledgements

We are very grateful to the ESRC's Consumer Data Research Centre for providing funding for this project. We are also extremely grateful to mySociety for allowing us to use the data described herein and being so open to collaboration.

Related research