Visual representation of safety in urban spaces: a tale of two neighbourhoods Open Access

Urban studies have addressed the role of the media with regard to the ways in which the production of news affects the practicing of urban life. The media, and newspapers in particular, are often used when identifying and providing evidence for urban research (Rodgers et al., 2009). Some studies focus on the media in terms of how it is theorised in urban politics (Ward, 2009), how newspapers play a role in urban growth (Maruani and Amit-Cohen, 2013; Raynor et al., 2017), or how various media in the city today can have an influence in the form of social/democratic commentary (Bridge, 2009). In this regard, this paper wishes to contribute to the field of urban research and urban planning by investigating the discourses that emerge from Swedish news media as a source of producing local knowledge. This is also in order to increase understanding of how media reports can influence urban planning, through the ways in which planners, citizens and local actors and participants perceive neighbourhoods and thereby are affected by news reports. This paper is part of a doctoral thesis investigating different perspectives on urban planning and justice – how environmental, procedural and recognitional justice are being considered and played out in planning projects. In this context, the focusing of media reports and images on urban neighbourhoods is seen as a way of revealing underlying estimations of neighbourhood characteristics, seemingly built on either actualities or prejudice. Such ideas are likely to play a role in urban planning and design.

In urban planning practice, local knowledge plays a central role when it comes to prioritisations and decisions for localisation of new investments. Local knowledge includes information related to local contexts or settings, including knowledge of specific characteristics, circumstances, events and relationships, as well as important understandings of their meaning (Corburn, 2003). Local discourses (i.e. how districts and neighbourhoods are talked about, and their reputation, especially in terms of safety and security) not only affect the prices of houses and apartments but also the ways in which politicians resonate and how planners act, with regard to changes in the urban landscape in the form of redevelopment and gentrification (Schwarze, 2022), even if this is not always openly declared. In planning literature, however, the links between local discourses and planning actions have not been thoroughly studied. Although this research field is admittedly complex, with difficulties in gaining an overview and the impossibility of ever being completed, there are some sources, such as newspapers, that reach citizens in the same way and to the same extent, possibly both reflecting and influencing predominant opinions. This paper aims to test the use of discourse analysis (DA) and content analysis (CA) on newspaper reporting of neighbourhood characteristics, in order to draw conclusions for how different neighbourhoods are described, including as representations of how these areas might be treated in urban planning.

There are two main categories of literature dealing with media and urban studies that have been found to be most relevant to the scope of this paper. By assigning meanings to a place, the media plays an important role in creating negative associations, giving specific names and symbolic defamation to particular urban spaces (Jensen and Christensen, 2012; Kearns et al., 2013; Nayak, 2019; Stjernborg et al., 2015). Stigmatisation is produced by the ways in which the media represents neighbourhoods by illustrating more negative images, such as crime and violence. Schwarze (2022) shows how newspapers reinforce stigma and racial boundaries by framing black communities as violent ‘no-go zones’. He states that ‘by using hyperbolic language, newspapers generate mental images of the community in the readers’ mind which work through graphic depictions of violence’ (Schwarze, 2022, p. 1424).

On the other hand, the media can also be used to promote identity, place branding and city development, which are discussed within the context of urban economy and urban growth (Avraham, 2000; Lavy et al., 2016; Paganoni, 2012). Rapid urbanisation and deindustrialisation enhance the justification of competitiveness in the domain discourse of city planning (Fainstein, 2010). One example is the city of Malmö, which is the third largest city in Sweden. The city has been transformed from an industrial city dominated by harbour and industry, to a ‘knowledge city’ by focusing on university and cultural production (Anderson, 2014; Holgersen, 2015). The transformation of the city through such development has succeeded in raising the image of the city as being more sustainable and green, but also resulted in increased segregation as part of the negative effect of gentrification (Holgersen and Malm, 2015). Malmö has been struggling with its image as a segregated city (Nylund, 2014; Stjernborg et al., 2015), with high rates of crime (Brå, 2019). In its annual report published in 2020, the Swedish National Council for Crime Prevention (Brottsförebyggande rådet) indicates that the rate of crime in Sweden has been increasing since 2006 (Brå, 2020). Swedish news media have also been reporting on an increase of unsolved crimes in Sweden – more specifically, in Malmö – in recent years (Expressen, 2015; Kvällsposten, 2012; Sydsvenskan, 2017). According to Malmö city, the sense of insecurity among Malmö residents remains high, although it has decreased in 2021 (71.8%) compared with 2020 (73.5%) (Malmö city, 2022). It is not surprising that places with social problems get more attention in news media, compared to other areas with fewer social problems (Lindgren, 2009). This paper seeks to explore how newspaper images and texts represent safety and security in urban spaces, specifically in the context of Malmö. The paper also aims to compare two selected neighbourhoods with different social status: one that is labelled as socially vulnerable (Bellevuegården), and one that is newly built and more privileged (Västra Hamnen).

According to social construction theory, individuals create their world within particular social contexts. The understanding of the world is created by history and culture, and it is therefore place-specific and historically contingent (Jørgensen and Phillips, 2002). Discourse as ‘a form of social action that plays a part in producing the social world – including knowledge, identities and social relations’ can inform the appearance of specific social patterns and processes (Jørgensen and Phillips, 2002, p. 5). For example, the photograph in Figure 1 was taken during the celebration of Sweden's National Day in 2008. The positive features of the discourse of Swedish nationalism can be elicited from the picture, which shows the Swedish flag located at almost the centre of the frame, being held by a blonde girl, looking towards the sky against a background of a huge number of blue and yellow-coloured balloons. Two more people can be seen at the bottom-right corner of the photograph, although they are in the background. The image was re-used in an article about Swedish values among the supporters of different political parties. The article says that supporters of the Sweden Democrats (the far-right party) had no problem naming Swedish values as values that are consistent with Swedish norms, traditions, rules and culture (Svenska Dagbladet, 2016). This example shows not only that text and conversations create reality, but also, as stated by Gillian Rose (2016), that visuality is the central part of the cultural construction of social life in contemporary Western societies (Rose, 2016), as visuality does not simply mean an image or object in itself, but is more about the perception and the meanings attributed to the images (Prosser, 2013).

Discourses are created and expressed through different sorts of resources, such as spoken and written texts, and visual and verbal images. As a result of the diverse forms in which a discourse can be articulated, the term ‘intertextuality’ becomes important in order to understand a discourse (Rose, 2016). Intertextuality points to the ways in which discourses are related and constructed across a combination of elements, which themselves draw from different discourses. For example, Norman Fairclough (1995) focuses on intertextuality as intertextual chains, which concern how one text links to another text (Fairclough, 1995). Rose (2016) includes images in the construction of intertextuality, and defines it as ‘the way that meanings of one discursive image or text depend not only on that one text or image, but also on the meanings carried by other images and texts’ (Rose, 2016, p. 188). Discourses and intertextuality are well demonstrated in the work of Schwarze (2022), which shows the role of newspapers in producing stigma concerning a black community in Chicago by depicting violence and crime (Schwarze, 2022). Another example is a study of the Danish neighbourhood Aalborg East, where Jensen and Christensen (2012) claim that ‘discourses related to ethnicity, age, gender and crime intersect. The categories at play, each a potential sign of trouble, intersect into a complex sign whose symbolic weight is more than the sum of its parts’ (Jensen and Christensen, 2012, p. 81). Here, it can be seen how the intertextual chains are being created, in the form of ethnicity, age, gender and crime.

At the same time, culture becomes a significant way of understanding how social processes, social identities and social changes are articulated in urban contexts. Stuart Hall (1997) explains culture as being ‘concerned with the production and exchange of meanings – ‘the giving and taking of meanings’ – between the members of a society or group’. Thus, culture depends on how individuals interpret their surroundings meaningfully and make sense of their world (Hall, 1997, p. 2). Adding to this argument, Rose (2016) claims that human-beings are surrounded by different kinds of visual technologies, such as photography, film, video, television programmes and newspaper pictures. In Western society, visual materials increasingly offer people a means to interact with one another on the basis of constructed representations of space (Rose, 2016).

What kind of social effects are caused by visual material, such as photographs? Rose (2016) first directs the attention to the question of how images visualise (or depict invisible) social differences, such as class, gender, race or sexuality, before examining how images are looked at. Visual methods examine images in order to unveil how social categories are constructed (Lynn and Lea, 2005; Prosser, 2013; Rose, 2016). For example, images of immigrants in news media might be more likely to depict them as a threat (Carvalho, 2010) or as deprived (Larsen and Dejgaard, 2013), or focus on their ‘otherness’ (Batziou, 2014). Another study examined the portrayal of demographics of poverty in five American news magazines, and found that significantly more African Americans were present in depictions of the poor than Whites, Hispanics or Asian Americans. The authors claim that, as a result of such representations of the poor in these magazines, many citizens overestimate the number of black people who are poor. They also found that images of Hispanics were absent in the data they analysed (Clawson and Trice, 2000), which points to what is ‘invisible’ (Rose, 2016, p. 17) or what is absent and not photographed (Lynn and Lea, 2005).

Another factor is not only how images look, but also how they are looked at. In other words, it is necessary to consider the importance of the role of the person looking at the image. Rose (2016) uses a quote from John Berger to point to a broader aspect of the way images are created, moved around and socially circulated: ‘we never look at just one thing; we are always looking at the relation between things and ourselves’ (Rose, 2016, p. 18). One example is what Lutz and Collins (1993) describe in their book Reading National Geographic as different types of gazes in the photograph and its social context. One of the gazes that they describe is the reader's gaze: ‘Independently of what the photographer or the caption writer may intend as the message of the photo, the reader can imagine something else’ (Lutz and Collins, 1993, p. 195). The reader's gaze is constructed individually, by their personal, cultural and political background (Lutz and Collins, 1993).

The purpose of this paper is to find out how discourses of safety and security emerge from reports in Swedish newspapers (local and national) concerning two neighbourhoods in Malmö: Bellevuegården and Västra Hamnen. The paper also investigates the ways in which media has been influential in the representation of security and safety in these neighbourhoods, thereby potentially affecting perceptions of how it is to live there. A visual discourse approach has been employed by combining CA and DA to examine the representation of crime and safety in Bellevuegården and Västra Hamnen. Drawing on an analysis of 323 reports, the paper will show specific ways in which social meanings are created and circulated through newspapers. One reason to include newspapers and no other sources is that, according to the latest report from Mediebarometern (Mediebarometern is an annual survey of the Swedish population's access to, and use of, different types of media (Mediebarometern, 2021).), 68% of the Swedish population aged 9–85 years old read a newspaper on a daily basis (Mediebarometern, 2021).

Prior research shows that news media can contribute to negative portrayals of disadvantaged neighbourhoods through biased representations of crime activities (Baranauskas, 2020; Halliday et al., 2020; Keene and Padilla, 2014) – specifically, as compared to well-off neighbourhoods. For example, consider the following stories from a well-known Swedish newspaper (Sydsvenskan). One article, published on 13 November 2016 (Figure 2, the Appendix), reports that a dead man was found in a car at the bottom of the water in Västra Hamnen – a well-off neighbourhood in Malmö. The other article reports another crime from Bellevuegården – a neighbourhood with lower income status and a higher population of people with immigrant background. The article reports that, on 10 June 2019 (Figure 3, the Appendix), a young man was shot in the head in a public place during the daytime. The way in which these two crimes are represented visually is significant. The images of Västra Hamnen show a long shot of a crane pulling a car out of the water. The images from Bellevuegården, on the other hand, all illustrate groups of police officers and police cars near the crime scene. Although these two examples describe the murder of a citizen in different circumstances, and both cases were as yet unsolved, they are made to look very different. While in the Bellevuegården case the police are active at the scene, no police is seen in the case of Västra Hamnen. Thus in Bellevuegården, an ongoing fight against crime is displayed (between the lines: between the police and the neighbourhood itself), while the image chosen in Västra Hamnen, there is no actual visual link made between the crime and the neighbourhood (it is more like a coincidence that a murdered man has been found there – the image itself does not show criminality). More specifically, they present two Malmö neighbourhoods as embodying very different levels of criminality and safety. Although the crime rates in these two districts are not equal, the focus of this paper is to examine how images can create specific meanings in their social contexts.

It is also argued that the consequences of media representations of different neighbourhoods have an effect on residents’ lives, raising the question of social justice in the city, specifically for vulnerable neighbourhoods. This means that a negative image of a stigmatised neighbourhood is not the only product of visual discourse created by news media but also results in an unwillingness on the part of the residents to become involved in their ‘loser’ community and/or that they want to move away as soon as they are able to (Lindgren, 2009, p. 93). Another consequence is, as argued by Castro and Lindbladh (2004) that living in a poor suburb increases the risk of vulnerability for young adults on both the individual and the community level. Different agencies such as media are counted as an important actor to create the hegemonic discourse of vulnerable areas (Castro and Lindbladh, 2004).

During the past two decades, the city of Malmö has undergone an extraordinary transformation, from an industrial city in deep crisis to a post-industrial ‘knowledge city’ (Holgersen, 2012). For example, there has been tremendous, increasing investment in two areas, (1) banks, real estate, business services and (2) education and research. Malmö city decided to revitalise the city by converting industrial land into housing for affluent citizens, and the construction of office buildings with the focus on the concept of sustainability (Holgersen, 2014). Holgersen and Malm (2015) use the term ‘green fix’ to explain how the local authority managed economic and ecological crises by introducing Malmö as a ‘green city’, which was a business strategy for the real estate market and also shown possible to export as commodity in itself (Holgersen and Malm, 2015).

Västra Hamnen is the product of a series of transformations of industrial land into an area for housing, offices and recreation. It is located on the coast to the west of the city's main harbour, on the site of a former shipyard. For some, the neighbourhood stands as proof of the fact that the transition from an industrial into a ‘knowledge city’ has changed the image of Malmö for the better: for example into a more attractive location in which to live, especially for the wealthier part of the population – essentially the tax payers (Holgersen, 2012; Holgersen and Malm, 2015). Among the characteristics of the area are that around 70% of the population in Västra Hamnen have post-secondary educational level (eftregymnasial), and their income is 30% more than the average for the city of Malmö (Malmö City, 2021). Despite the benefits of urban greening, the green fix strategy creates socio-spatial inequalities (Anguelovski et al., 2018) (Figure 4).

Increasing segregation, and disparities in income and poverty between ethnic groups, are results of both demographic changes, due to globalisation and a high rate of immigration (Nylund, 2014; Stjernborg et al., 2015), and green strategies (Holgersen and Malm, 2015). Bellevuegården is one of the districts in Malmö that has been labelled as a ‘vulnerable area’ (The Swedish Police has established a list of ‘vulnerable areas’, in order to allocate extra resources and protocols to deal with their crime rate. According to the report by the Swedish Police, there are three types of vulnerable area: (a) a vulnerable area presents a problem that primarily requires police presence and police efforts to deal with existing problems; (b) a particularly vulnerable area presents a problem that, to a large extent, requires joint efforts from several societal actors in order to address co-existing problems and (c) a risk area is an area that meets all the criteria for a vulnerable area but does not really meet the criteria that characterise a particularly vulnerable area. However, the situation in a risk area is so alarming that there is an imminent risk that the area could become particularly vulnerable unless adequate measures are put in place (Swedish Police, 2015)), and was built in 1974 during the Million Programme era (The Million Programme was a national housing programme that aimed to build approximately 1 million homes during the period 1965–1974. The first decades of the post-war era in Sweden, with rapid urbanisation and growing prosperity, led to a housing shortage. To rectify this, the government (formed by the Social Democratic Party) decided to build 1 million new apartments. This ambitious project was intended to provide housing for middle-class workers around the bigger cities. About halfway through the Million Programme, the housing deficit became a housing surplus, partly due to the rapid expansion of the housing stock, and partly due to the economic crisis (Grundström and Molina, 2016; Hall and Viden, 2005). The project was not as well-received as the government had expected, and by the end of the 1960s, the project was already being criticised due to what some people perceived as uniform and poor architecture (Hall and Viden, 2005). The areas in which the scheme's housing had been built became unattractive places in which to live, and many apartments remained empty. To solve the vacancy problem, many municipalities accommodated large immigrant populations with lower social status who did not have any other choice (Mack, 2019). For this reason, the Million Programme neighbourhoods have become associated with ethnic segregation and poverty, and Bellevuegården (as our focus area in this paper) is no exception). According to the Swedish police, Bellevuegården has been experiencing social problems since 2015, together with other vulnerable areas in Malmö and other parts of Sweden (Swedish Police, 2017). According to Malmö Municipality's website, the residents of Bellevuegården have low economic status, and a higher level of unemployment (40%) and lower level of education (compared to Västra Hamnen) (Malmö City, 2022), and, importantly, the local community has been influenced by criminal activities (Police, 2017). The latest report published by the Swedish Police concerning vulnerable areas stated that the situation in Bellevuegården has improved, but there is still a risk that the improvement trends could be reversed (Swedish Police, 2019). What is specific for Bellevuegården is its elderly population, with up to 25% of residents being more than 65 years old, and the proportion of residents with a foreign background is around 45% (Malmö City, 2022). The main differences between Västra Hamnen and Bellevuegården concern socio-economic status, level of education, income and ethnic background.

To study the effect of visual representation, two districts in Malmö were chosen: Bellevuegården and Västra Hamnen. The data were collected among the relevant reports that concern about safety and security in these two neighbourhoods from Swedish media database ‘Retriever’ (Retriever is a digital archive hosted by Retriever Sverige AB, covering printed press, web content, radio and television coverage, for all Scandinavian countries and some European (Swedish, Norwegian, Danish, Finnish, Finland-Swedish, German, French, Spanish and English) dating back to the 1980s. (https://www.retriever.se/product/mediearkivet/)). The reports were browsed from 22 April 1998 (the first data which were available in the archive started on that date) to 01 January 2020 by applying the search strings in four groups: (a) ‘Bellevuegården’ AND ‘Malmö’ AND ‘safety’ (in the actual Swedish searches, ‘trygghet’ was used which can be translated into English as a feeling of safety/security, and ‘otrygghet’ as insecurity and feeling unsafe (Rodenstedt, 2014, p. 49)), (b) ‘Bellevuegården’ AND ‘Malmö’ AND ‘crime’, (c) ‘Västra Hamnen’ AND ‘Malmö’ AND ‘safety’ and (d) ‘Västra Hamnen’ AND ‘Malmö’ AND ‘crime’. Only local and national newspapers were included among other possible sources, such as radio and TV programmes. This selection was made for the methodological reason that the layouts of Swedish newspapers are quite similar in terms of images, titles and columns, which makes it easier to collect and interpret the data. This paper includes local and national newspapers in order to understand the production of local discourse at the city level and their relations to a broader perspective of national level. In total, there were 192 reports from Västra Hamnen and 131 reports from Bellevuegården that contained any of the words ‘safety’ or ‘crime’ in both areas, either in the text, the title or the picture captions. Both relevant articles (It means those reports that are directly about each neighbourhood.) and indirectly relevant articles (It means that the report is about something else other than Västra Hamnen or Bellevuegården (e.g. reports that just mention these neighbourhoods somewhere in the text or in the picture caption).) were kept.

CA and DA were employed and combined to analyse the data. According to Rose (2016), CA is an appropriate method for dealing with a large number of images. DA was used, which is an appropriate method for comprehending intertextuality – that is, how different texts interrelate with each other (Rose, 2016). Although CA and DA come from a very different philosophical base (Hardy et al., 2004), CA and DA were combined for the purpose of this paper in order to understand not only the effect of the visual images but also to add extracted meanings from the discourses. CA was used in order to see what was illustrated by the newspapers, and how the contents of the images can be described. DA was used to look more closely at the images – to identify what is visible as well as what is invisible, as Rose (2016) mentions that ‘invisibility can have just as powerful effects as visibility’ (Rose, 2016, p. 213). CA is based on four steps, of which the first is to find the images in accordance with research questions. This paper considered the images that appeared in newspaper reports that consist of ‘safety’, and ‘crime’ as keywords in both neighbourhoods. The second step was to create a group of categories in such a way that each aspect of the image would belong to one category. The categories must not overlap, and they should produce interesting coherent information. The third step was to code the categories in such a way that the coding process would be replicable. The fourth and final step was to analyse the results, which could be done in various ways, but the simplest way was to count the frequencies (Rose, 2016).

The aim of using DA was to explore how images that form a particular view of the social world can be constructed. If the fact is accepted that discourses are produced socially, then this type of DA is especially interested in the image itself, social production and effects of discourses (Rose, 2016). However to understand the meaning of a discourse and the social production of it, it is necessary to know the importance of intertextuality, which Fairclough explains as ‘how an individual text draws on elements and discourses of other texts’ (Jørgensen and Phillips, 2002, p. 7). This means that all spoken and written statements necessarily draw on and incorporate earlier (and parallel) spoken and written statements (Jørgensen and Phillips, 2002). Intertextuality therefore requires a certain breadth of source material.

The material was first analysed with CA, using the four-step technique suggested by Rose (2016): (a) selecting relevant images for research questions, collecting all the reports from each neighbourhood, eliminating the reports without any image; (b) creating a list of categories for coding; (c) coding the images and (d) interpreting the results (Rose, 2016). Then DA was used for all the reports, regardless of whether there was an image or not, to investigate how news and reports assign specific meanings to Bellevuegården and Västra Hamnen. The process of analysis started by reading and re-reading the texts and headlines that contained the words Bellevuegården or Västra Hamnen. Two categories were recognised. First, the news was about Bellevuegården or Västra Hamnen, and that something had happened there. Second, the news topic was about something else, and Bellevuegården or Västra Hamnen were just mentioned somewhere in the text. For the first category, after examining the text and image, a short description was written of what it was about, and examined which objects were represented in them. For the second group (To clarify how the second category was defined, here is one example of a report from Sydsvenskan published on 11 August 2019 about Seved – a neighbourhood in Malmö with social problems. The title is ‘Seved is on the right track – but the fear remains’, and it explains that, although the engagement of different groups, such as police, municipality and stakeholders, has improved the level of safety in the neighbourhood, there is still evidence of shootings in public places. At the end of the report, Bellevuegården was described as ‘characterised by open drug dealing, dissatisfaction with society, acts of violence that harm third parties’ (Sydsvenskan, 2019b). In this category, Bellevuegården and/or Västra Hamnen are stated in the reports were implicitly assigned to either negative or positive meanings depending on the whole story of the report.), the same process was followed, as well as assessing where the name of Bellevuegården or Västra Hamnen was mentioned, and how it related to the main topic of the article. The next step was to identify a central theme that emerged from the texts, headlines and images.

The paper studied 131 reports from Bellevuegården and 192 reports from Västra Hamnen, both directly and indirectly concerned with safety and security. Among these, 28 reports from Bellevuegården and 53 reports from Västra Hamnen contained images. It also included 38 reports from Bellevuegården and 50 reports from Västra Hamnen that featured indirect news but contained images (one image or more) (Table 1).

Figures 5 and 6 show the content of images from directly relevant news of Bellevuegården and Västra Hamnen. By comparing these two diagrams, it can be seen that around 80% of the focus of the news from Bellevuegården is on the crime (by showing images of police and crime scenes), while only 45% of the reports from Västra Hamnen focuses on crime and crime scenes. This is quite significant, considering the number of reports from Bellevuegården compared to those from Västra Hamnen (Table 1). There are even differences in the ways in which the images showing crime scenes and police officers have been visualised in these two areas. The photographs published from Västra Hamnen that depict serious crime are mostly taken from a distance, which makes it hard to recognise police officers, for example. By contrast, in Bellevuegården, the viewer can often discern the faces of police officers, which suggests a nearness or proximity to the threat of crime (and thus the need for security) (e.g. Figures 2 and 3, the Appendix). One interpretation of these different depictions of police in the two neighbourhoods could be to emphasise the level of crime – for example, crimes in Bellevuegården are more serious than in Västra Hamnen, due to the higher number of police officers in Bellevuegården. Here, it is important to emphasise that, depending on socio-cultural background, the interpretation of these images can be different – for example, the police presence could create a sense of reassurance and protection, or, on the other hand, it could create a feeling of fear and insecurity due to the proximity to crime. There is not so much difference in depictions of people in the reports from both neighbourhoods, but visualisation of outdoor environments and city development is much more in focus in Västra Hamnen.

Examples of indirectly relevant news from Bellevuegården and Västra Hamnen are shown in Figures 7 and 8. Both direct and indirect news from Bellevuegården demonstrate that, in total, more than 80% of images in the total number of reports from Bellevuegården are about crime (percentage of photographs depicting crime scenes or crime-related scenes, such as police cars, officers and ambulances). In Västra Hamnen, the equivalent figure is 50%. It is important to note that the reports in this group were not specifically about these neighbourhoods – the main topics were actually about something else. The remaining images depicted more people with immigrant backgrounds and politicians in Bellevuegården. However, there were also a variety of pictures showing people such as native Swedish together with immigrants in outdoor environments (e.g. beaches) in Västra Hamnen. There were also images depicting the former mayor of Malmö city, Illmar Reepalu, who had a great impact on the development of Västra Hamnen (Holgersen, 2014).

Tables 2 and 3 demonstrate the main themes that emerged from the DA for Bellevuegården and Västra Hamnen. Reports from both areas show that the main theme is crime, although the number of news reports that were from Bellevuegården was more than twice those from Västra Hamnen. The results show that, on the one hand, the media is producing a discourse of insecurity through crime-related news from Bellevuegården. On the other, the results show the breadth of news about crime prevention as densification development in Bellevuegården, and housing companies’ special programmes for the area. As Table 2 shows, the majority of the reports focus on crime incidents in Bellevuegården, which explicitly and implicitly concern issues of danger and insecurity in the area. Accordingly, a few reports described how Malmö Municipality has presented the new densification development in Bellevuegården by concentrating on the crime prevention aspect of the design. This is one way to argue for densification of the neighbourhood. At the same time, there were some reports about residents’ protests against the densification. Moreover, housing companies, such as Malmö Kommuns Bolag (MKB), ran projects with the aim of making the neighbourhood safe. In contrast, there was a variety of topics about Västra Hamnen (e.g. sustainability, education, regional growth, famous example of building, public space structure). It is worth mentioning that Västra Hamnen was described using adjectives such as attractive, excellent, sustainable, safe place, expensive, luxury, rich, diverse and dense. This is what Rose (2016) emphasises as how specific visuality can make certain things visible in a special way, and certain other things invisible (Rose, 2016). In this case, what is made most visible from Västra Hamnen is the idea of an area that is an attractive and nice place, both for residents and for business.

Table 4 presents the results for indirectly relevant news concerning Bellevuegården. It shows two main themes: first, the reports about serious crime in Malmö city or elsewhere in Sweden, and second, the reports about vulnerable areas in Sweden. The incidents of serious crime, such as shootings, murders and lethal violence, repeatedly mention previous incidents from Bellevuegården as well as other places. Moreover, there are follow-up reports about the suspects, police investigations and, more importantly, reports about unsolved crime in Malmö. There is a special phenomenon for reports about vulnerable areas in Swedish newspapers. Whenever there is news from a vulnerable area in Sweden, there is a separate column that defines vulnerable areas and lists places that belong to these groups. In this way, the reader gets the extra information about for example, Bellevuegården, which is on the list of vulnerable areas, even if the report is actually about somewhere else. Furthermore, there were a few reports focusing on crime prevention, such as the installation of cameras or various projects carried out by housing companies or other organisations.

Table 5 shows the indirect reports concerning Västra Hamnen, which illustrate mainly positive perceptions. Most of the reports that mention Västra Hamnen describe it as an attractive, expensive, safe and sustainable location. Västra Hamnen is cited as a good example for other new developments, and is also compared positively to the most problematic areas in Malmö, such as Rosengård and others. Although Västra Hamnen was mentioned in the crime-related reports from Malmö/Sweden, there are more reports that show Västra Hamnen together with the Öresund Bridge, the city tunnel and Malmö University as being part of the positive side of Malmö.

Throughout this paper, possible discourses of safety and security have been explored, as illustrated in newspaper representations of two neighbourhoods in the Swedish city of Malmö: Bellevuegården and Västra Hamnen. Unsurprisingly, reflecting the earlier descriptions of the two neighbourhoods, the overall results show that the image of Bellevuegården regarding safety is negative, whereas the image of Västra Hamnen is of an attractive, lively and safe place to live. This is aligned with previous research which found that media tends to cover news from disadvantaged neighbourhoods disproportionately (Baranauskas, 2020; Halliday et al., 2020; Stjernborg et al., 2015). Bellevuegården and Västra Hamnen differ in physical and social aspects. Bellevuegården belongs to the Million Programme era of the 1970s, while Västra Hamnen is a newly developed area which, by having access to the waterfront, is much more attractive. Their social environments also vary, with Bellevuegården's residents having lower social status, lower income and a higher level of unemployment, and, more importantly, it has been listed as a vulnerable area of Malmö. Västra Hamnen, on the other hand, is a place for wealthier residents with higher social status. The negative image of Bellevuegården and the positive depiction of Västra Hamnen, as explicitly shown in this paper, came as no surprise. The contribution of this study is that the current image of each neighbourhood is the result of a more complex background, and it would be suggested that the concept of intertextuality can explain this complexity by showing how different discourses interrelate and overlap with each other.

As stated earlier, multi-storey Million Programme buildings have been associated with ethnicity and migration, poverty, low socio-economic status and social problems such as drug dealing, rioting, violence and even murder (Backvall, 2019; Mack, 2021; Wiard and Pereira, 2019), which together created a discourse about the Million Programme. It is argued that presenting a list of the names of vulnerable areas in a separate column can draw attention in a visual sense. One might not need to read the whole story to understand what has happened. Instead, merely by looking at the title, the separate column and the images, one can associate them (title, images etc.) with Million Programme housing, vulnerable areas and social problems. Figure 9 (the Appendix) is an example that shows how the combination of text and image can create a discourse of the perceived unsafe situation in vulnerable neighbourhoods. Here, the focus is on the interrelationships between texts/images and the other texts/images as intertextuality.

Rose (2016) defines intertextuality as how the meaning of one discursive image/text relates to the meaning of other images/texts. Such intertextuality in the images/texts can be seen from Bellevuegården, for example, where the text is not about Bellevuegården but mentions Bellevuegården's status as a vulnerable area. This connects the discourse of the Million Programme to Bellevuegården, while involving associated problems such as segregation, poverty, unemployment and violence. Figure 9 (the Appendix) shows an example from Rosengård, which is a particularly vulnerable neighbourhood in Malmö built during the Million Programme era. Reading this report, one associates and interrelates problems with Million Programme residential areas in general. When the article states which areas are still vulnerable neighbourhoods, it reminds the reader that Bellevuegården is still not safe. This is also aligned with the quote from John Berger, which says that ‘we never look at just one thing; we are always looking at the relation between things and ourselves’ (Rose, 2016, p. 18). In this case, the interpretations could differ if, for example, one had an immigrant background and lived in such an environment or if the person was Swedish Native who had never visited such a place. According to the viewer's personal, cultural and political background the images are looked at (Lutz and Collins, 1993; Rose, 2016). This does not mean that both interpreters would not know, in a superficial sense, that a ‘vulnerable area’ is tainted with negative characteristics. What it means is for example that while a person living in such a neighbourhood would have a more diverse and broad conception of the place, more mixed with positive feelings and sense of place, the stranger is more likely to believe the media reports in a literal sense. Since those unfamiliar with vulnerable areas dominate the representatives and authorities of the society, this is likely to cause a bias in terms of prejudice, in questions and assignments associated with these areas.

One of the categories of reports from Bellevuegården concerns the ongoing densification project in the neighbourhood. These reports may sound neutral or even positive, but Jensen and Christensen (2012) call them ‘paradoxical stories’, as the overall approach of these reports is often positive, but they still imply a negative description of the area (Jensen and Christensen, 2012, p. 80). In the case of Bellevuegården, it can be seen that the argument for the development and densification project is to make the area safe and secure. This is in alignment with the literature, which discusses how the negative image produced by the media serves capitalism by arguing that such stigmatisation results in the devaluation of an urban area, which means that redevelopment and revitalisation are needed in order to ‘revive’ the area (Parisi and Holcornb, 1994; Schwarze, 2022; Willer, 2021).

To clarify the intertextuality between Bellevuegården and other vulnerable areas and the discourse of the Million Programme, here is a concrete example from Sydsvenskan (Sydsvenskan, 2019a). An article about Rosengård has the headline ‘A neighbourhood to love – or to be afraid of’ (‘En stadsdel att älska – eller vara rädd för’) (Figure 9, the Appendix). The report contains four images, the biggest of which shows two women with an immigrant background sitting on a bench in the shopping centre in Rosengård and looking happy. The caption reads: ‘I have heard many bad things about Rosengård, so it is good to highlight the positive aspects here: there is a warmth.’ (‘Jag har hört många dåliga saker om Rosengård och därför är det bra att lyfta upp det bra här: Det finns en värme.’) The next image shows a Swedish-looking woman with smile on her face, saying that ‘You don't need to go away on holiday. There is a lot of culture – when you come here to shop, there is a completely different range. The spices, for example – if you don't know what to do with them, you get given a recipe for free.’ (‘Ja, man behöver inte åka på semester. Det är mycket kultur, när man går upp här och handlar är det ett helt annat utbud. Kryddorna till exempel – när man inte vet vad man ska göra med dem får man ett recept på köpet.’) Although these images focus on a positive aspect of this vulnerable area, the text still implicitly focuses it being unsafe, including a picture showing fire and explosions. There is also a special column that provides information such as the rate of employment (36%), the proportion of the population with a foreign background (90%), and the proportion of families receiving financial support (37.7%). In the newspaper article, Bellevuegården is mentioned as a vulnerable area. All of this information, from the images and the text, combines to create a mental image with specific meanings, which by means of intertextuality one relates to their previous knowledge. It has been argued that this process empowers and augments the discourse of the Million Programme. This is an example of how a basically positive news article can be given another tone, thereby contributing to a more negative discourse that was not central to the content of the article. It can therefore be argued that it is very hard to change a negative image. This is partly due to the superficial and judgemental ‘positive’ reports, and partly due to culturally biased perspectives on safety and security, most of which lack an understanding of spatially relational differences between how the neighbourhood is perceived in comparison with other urban areas (Listerborn, 2014).

The interrelationships and intertextuality between each neighbourhood (Bellevuegården and Västra Hamnen) and the city of Malmö as a whole have been recognised. Swedish newspapers play an important role in the creation of the discourse of insecurity in certain parts of the city, and a discourse of safety and security in the other parts. Such discourses are part of the construction of a city that is segregated by class, social status and ethnic background (Grundström, 2019). On the one hand, there is an image of a dangerous city (Martin, 2000; Schclarek Mulinari, 2017; Stjernborg et al., 2015), where Bellevuegården is part of it as one of the vulnerable areas. On the other, the attractive, exclusive and more privileged area of Västra Hamnen would improve the negative image of Malmö. However, Västra Hamnen is not totally safe from criminals, as the results show, but there is quite a strong emphasis on the high quality of life in Västra Hamnen in both direct and indirect news reports.

Västra Hamnen as positive side of Malmö constructed a segregated city, contradicts the vision of Malmö city, which aims to ensure social sustainability and to reduce inequality among the residents (Stigendal and Östergren, 2013). It has been claimed that the process of dividing the city of Malmö into safe and unsafe areas, is partially caused by the impact of indirectly relevant newspaper articles concerning both neighbourhoods. For example, more than 30% of articles describe Västra Hamnen in one or two sentences, using descriptive adjectives such as attractive, exclusive, successful and sustainable, or describe it as a good example for future developments in other cities. In addition, around 15% of the articles compare Västra Hamnen to other, more vulnerable, areas in Malmö, which increases the focus on the problems in those areas, as well as the discourse of the Million Programme that emphasises segregation. This trend of increased social inequalities is in direct conflict with the ambition for sustainable city development (Stigendal and Östergren, 2013).

This paper explains how representations by news media explicitly and implicitly influence the image of Bellevuegården and Västra Hamnen. What is important to consider, are the consequences of such discourses and how they can affect residents’ lives. The redevelopment of Västra Hamnen results in more inequality and social injustice in the city of Malmö, which Harvey (2019) criticises as a capitalist strategy to increase the flow of capital through development and the promotion of tourism (Harvey, 2019). Holgersen and Malm (2015) argue that a neighbourhood can become favourable on the basis of its ability to attract more business and higher property values (Holgersen and Malm, 2015). The issue of investment being unevenly distributed in Malmö has also been discussed in a newspaper article (Sydsvenskan, 2021). It explains that Malmö Municipality invests three times more in the rich areas than in the poorer ones, which contributes to increased inequality and segregation. Moreover, Mitchell (2017) argues for a relational approach to landscape and urbanism. He explains that landscape is relational, and this relationality enables individuals to understand the construction of such segregation which means that the development of landscape of privilege and marginalised are dialectically entwined with each other (Mitchell, 2017).

Furthermore, this is not only about labelling places with ‘good’ or ‘bad’ meanings; more importantly, it is about how these meanings come to play an active role in peoples’ lives. Stigmatised neighbourhoods and their inhabitants experience a range of negative consequences on their health and well-being by being exposed to greater stress (Halliday et al., 2020; Kelaher et al., 2010; Tran et al., 2020). There is limited access to resources and opportunities (Keene and Padilla, 2014) – what Andersson describes as ‘places of few opportunities’ (Andersson, 2016). For example, children from disadvantaged neighbourhoods experience failure in their education at school, with consequent employment in low-paid jobs (Gulczyńska, 2018).

Discourses of safety and crime in Swedish cities have increasingly become part of the political debate in recent years, and this accelerates the process of segregation. In the Swedish context, Sager and Mulinari (2018) argue that the discourses of safety and crime produced by nationalist and racist agendas create socio-spatial political and economic patterns in the city, which raise the question of ‘safety for whom?’ The authors argue that, within ethnonationalist ideologies, the notion of ‘safety’ means to protect and care for white women, and the logic of femonationalism considers gender equality to be a national cultural value, with ‘the others’ (both male and female immigrants) threatening this gender equality (Sager and Mulinari, 2018). The socio-spatial political and economic status can explain why the residents of vulnerable and unsafe neighbourhoods feel at home, even though the news reports about crime concentrate on such areas more than on other parts of the city. In Geographies of the Veil, Listerborn (2014) examines the meaning of violence for Muslim women in Malmö, which develops in relation to place and space. The narratives from Muslim women reveal that socio-economic segregation of the city can be overlapped with socio-spatial order in the public space, as some areas are defined in terms of being Swedish or immigrant areas (Listerborn, 2014).

There is quite a substantial body of literature on the effects of alternative approaches to master planning and neighbourhood design, with regard to behaviour, feelings of comfort and life quality, everyday functions and sense of place (Douglas, 2010; Mack, 2021; Romice et al., 2020). This paper does not counteract or contradict these findings, but argues that the media plays an influential role with regard to prejudice and expectations of neighbourhoods. Media narratives concerning neighbourhoods can therefore affect societal interventions, such as the allocation of resources, urban planning, infrastructure and design. As segregation has shown adverse effects on urban life quality and possibilities (Andersson, 2016; Keene and Padilla, 2014), it is important both to detect biased and pejorative reporting about neighbourhoods, and to find ways to counteract prejudice and support environmental, procedural and recognitional justice in urban planning and design. Suggestions for further research include critical studies of segregated cities (on a comprehensive planning level), considering alternative spatial-demographic divisions, as well as methodological development for participatory neighbourhood design and management.

It has been concluded that in order to promote social development with equal upbringing and living conditions and good life chances for everyone, it is especially important for urban planners to consider the role of the media in the creation of such discourses. The Swedish government's strategy for decreasing segregation states that segregation is a relational phenomenon that includes an entire city or region. It is in alignment with the discussion of this paper, and maintains that segregation is determined by the relationships between, for example, two or more residential areas (Regeringen, 2022). It is therefore not only areas with socio-economic challenges that ‘are’ segregated, but also that developments in different residential areas affect each other. In this regard, the opportunities to reduce the counteracting residential segregation are affected by handing over the responsibility for the design of cities to urban planners at municipalities.

End note

The title of this paper is clearly a paraphrase of Charles Dickens’ historic novel ‘A Tale of Two Cities’. Even if neither purpose nor scope is shared with this novel, its famous opening lines could apply to a dystopic view of a segregated city:

The author gratefully acknowledges the support of Dr. Max Ritts and Dr. Gunilla Lindholm who helped them throughout the writing of this paper. The author also extends his appreciation to Dr. Christine Haaland, Dr. Sanna Stålhammar and Frederik Aagaard Hagemann for their insight and feedback on an early version of this paper. This paper was funded by VIVA-PLAN Formas project (grant 2018-00175).

See Figure 2, Figure 3 and Figure 9.