Introduction

Public libraries play a central role in our society. The resources and services they offer create opportunities for learning, support literacy and education and help shape the new ideas and perspectives that are central to a creative and innovative society (White 2012). Public libraries are also multifunctional spaces. They have become a meeting place where people come to borrow books, read, study, and use computers but also to socialise. Although some public libraries can be hard hit by crime, we know very little about their security challenges and whether they are similar to those found in other semi-public spaces such as shopping centres or transport hubs.

Public libraries are interesting objects of study given their particular design, but also because visitors carry out different activities at various activity nodes (Brantingham and Brantingham 1993), which may sometimes create necessary conditions for safety-related incidents such as shouting, sleeping, eating or smoking on the premises. Public libraries, like many other risky facilities (Clarke and Eck 2007), are also interesting settings from a theoretical perspective because they constitute good examples of why certain micro-places attract crime. Micro-places are important as a unit of analysis because a large number of crimes tend to be concentrated in a small proportion of these settings (Weisburd 2015). The importance of analysing micro-places such as bars (Sherman et al. 1989) and street segments (e.g. Favarin 2018; Weisburd 2015) has long been reported in the international criminological literature (Curman et al. 2014; Taylor 1997), but relatively few studies have focused on the quality of indoor environments (see e.g. the work in jails and shopping centres carried out by Rengert and Ratcliffe 2005; Ceccato et al. 2018).

Drawing from environmental criminology theory in general, and the principles of Crime Prevention Through Environmental Design (CPTED) in particular, this study proposes and tests a conceptual model to investigate the spatiotemporal nature of crime and other safety-related incidents in public libraries. The aim of this study is to contribute to a better understanding of these internal micro-places in public libraries and is aided by empirical data from two public libraries in Stockholm, the capital of Sweden. Using a database of incidents recorded by library personnel from 2017 to 2020, crime and incidents of public disturbance are linked and mapped using a three-dimensional model. The methodology also includes fieldwork inspection and interviews with personnel.

We will use the term library for simplicity from now on as a synonym for public library. The focus is on public libraries as opposed to university libraries, as they may differ in terms of ownership and management. Public libraries may be part of community centres in Sweden but rarely do they combine functions with those found in community centres. We compare two public libraries to reveal potential differences that can account for the differences in risk, thereby providing important indications for crime prevention. Stadsbiblioteket is the largest library in Stockholm and located centrally, while Högdalen library is a ‘metro-library’, which means that it is located near or at a metro station in the periphery of the city. Beyond location and size, the two libraries also differ in layout, design, and activities that they promote, which makes it ideal to test whether the conceptual model can be used in different libraries in different neighbourhoods and city contexts.

This case study aims to understand the criminogenic characteristics of public libraries as risky facilities (Clarke and Eck 2007). The study combines multi-data sources and makes use of CAD—Computer-Aided Design software—in environmental criminology for visualisation of incident locations and concentrations in libraries in multistorey buildings. In addition, the research contributes to the international literature by illustrating cases in a Nordic European context, which has so far been dominated by cases from North America or the United Kingdom (an exception is e.g. Adewuyi and Adekanye 2011). In addition, it is important to keep in mind that with Covid-19 restrictions, the number of visitors decreased in 2020 (despite the fact that the libraries in Stockholm were open to the public as usual, as were the few public establishments providing access to public computers and copy machines), indirectly affecting the criminogenic conditions of all public libraries. There were a few incidents with the personnel related to Covid-19 restrictions which were excluded from the dataset.

The article is divided into eight parts. First, we discuss the theory underlying the model, then we propose the conceptual model for public libraries. The case study is framed, followed by a description of the methods. Then, the spatiotemporal nature of crime and public disturbance in the libraries is presented. The article finishes by bringing together the evidence from a fieldwork inspection, interviews with personnel and evaluation of maps, as well as suggestions for improving the library environment.

Theoretical background

Crime and acts of disturbance in public libraries

Public libraries, like many other facilities, are places with specific public or private functions (Clarke and Eck 2007) where incidents of public disturbance (e.g. shouting, sleeping, eating) as well as crime are concentrated. According to Clarke and Bichler-Robertson (1998), there is no single reason why facilities vary in risk, but it is important to determine which reasons are in operation in each particular case, as the facilities vary in size, design and layout, location and management.

We adopt several complementary theoretical perspectives as a basis for our understanding of public libraries as risky facilities. Facilities are not equally affected by crime and incidents of public disturbance, partially because the design and diverse use of their built environment can affect these incidents (Crowe 2000). These principles are often used under the umbrella term Crime Prevention Through Environmental Design (CPTED) (Jeffery 1971). The value of the CPTED approach resides in an attempt to gain a better understanding of the effect of the environment on individuals’ behaviour, either as offenders or as crime targets in libraries, but note that the literature has also shown evidence of the importance of CPTED to the study of fear of crime (Abdullah et al. 2013; Cozens and Sun 2019). Individuals’ behaviour is not static; it follows particular temporal rhythms that affect crime levels, as crime is dependent on the confluence at a particular place of a motivated offender, a target and absence of guardians (Cohen and Felson 1979).

Libraries contain a variety of ‘activity nodes’ (Brantingham and Brantingham 1993) at particular micro-places, such as receptions, reading areas and cafés, where visitors undertake different activities. Although these micro-places may have, by their nature, a predisposition to create the conditions necessary for crime, the quality of the environment may prevent it from happening by promoting natural surveillance, territoriality or having good management practices in place (Clarke 1983; Jacobs 1961; Newman 1972).

Poor management affects crime and incidents of public disturbance (Eck 2019). Previous research shows that few rules, lax enforcement, poorly trained personnel, etc., can encourage individuals to commit crimes, and get away without being noticed (Clarke and Eck 2007; Linning and Eck 2021). Multiple actors (handlers, managers, guardians) who can exercise social control have been shown to be key in preventing crime (Felson 2006; Linning and Eck 2021), but it is unclear what their role is in different circumstances. Yet, some facilities create so many crime opportunities and opportunities for public disturbance that they stand out as crime generators. Crime generators are places to which large numbers of people are attracted for reasons unrelated to criminal motivation; as such, they provide large numbers of opportunities for offenders and targets to come together in time and place to produce crime or disorder (Brantingham and Brantingham 1995). In criminogenic neighbourhoods, libraries, like any public facility, may absorb crime and incidents of public disturbance from the surrounding areas; in other cases, they may radiate crime to the surrounding areas (Bowers 2014).

In the next section, we discuss how previous research has reported on how these factors affect safety in libraries in more detail and the ways in which situational crime prevention has been applied in these types of facilities.

Crime, acts of public disturbance and safety in libraries

Studies on university libraries dominate the international literature on security in libraries (e.g. Adewuyi and Adekanye 2011; Henrich and Stoddart 2016). Although these facilities may not be the same as public libraries, they share commonalities as risky facilities. In the next sections, we discuss some of findings from this research without making any distinction between them. Note that this international research is a mixture of empirical evidence as well as checklists and recommendations, which may not have been tested.

Libraries can attract many types of crimes, from thefts (Adewuyi and Adekanye 2011; Carey 2008; Cromwell et al. 2008; Simmons 2018), assaults and threats against library staff or patrons (Adewuyi and Adekanye 2011; Carey 2008; Cromwell et al. 2008) to property damage, vandalism and other incidents of public disturbance (Carey 2008; Cromwell et al. 2008; Henrich and Stoddart 2016). Other relevant safety concerns in libraries are reported in the international literature, such as inappropriate use of library computers (such as to access pornography), inappropriate use of library space (sleeping in library) by the homeless and the mentally ill (Cromwell et al. 2008), lax regulations that allow people to carry firearms in college (Henrich and Stoddart 2016), or concerns about valuable collections and the safety of property in the libraries (Carey 2008; Cromwell et al. 2008).

Design and layout have an important role to play in the safety conditions of libraries. The international literature shows that crime events occur in hidden places, such as lift access points, stairwells, hallways (Henrich and Stoddart 2016), exits from emergency exit stairs and entrapped and enclosed rooms (Carey 2008; Henrich and Stoddart 2016). Areas with poor natural surveillance, such as behind and between bookshelves, have great potential for crime (Cromwell et al. 2008), as well as dark areas and corners that lack proper illumination (Carey 2008; Cromwell et al. 2008; Henrich and Stoddart 2016). Some areas are criminogenic because of the activities they attract. For example, areas of increased safety concerns are settings where the children’s and teens’ periodicals are located but also the adult fiction reading room, were the most diverse clientele may be found (Carey 2008). Rooms with valuable collections of books also known as CRAVED products, such as journals or newspapers in popular demand as well as personal lap top computers, DVD players, and cell phones (CRAVED - Concealable, Removable, Available, Valuable, Enjoyable, and Disposable) (Carey 2008; Cromwell et al. 2008; Smith 2018).

Crime and acts of public disturbance in libraries do not occur at random, either in space or time. According to Simmons (2018), criminal activities related to theft of a patron’s property occur mainly when they leave their belongings unattended, which can be easily prevented: almost all the thefts occurred during peak visitors’ hours during the afternoon, from October (mid-term) through December (finals), and the time in the day is mainly between 3 p.m. to 5 p.m. Regardless of opening hours, criminal events tend to occur mostly during people’s ‘working hours’ (Carey 2008; Cromwell et al. 2008; Henrich and Stoddart 2016), and the authors suggest more strict safety measures at these specific periods of time.

Crime prevention and safety interventions in libraries

There are many ways to prevent crime and promote safety in libraries. Opportunity theory relates changes in the nature and amount of crime to changes in the so-called opportunity structure for crime. Any increase in the difficulty or the risks of crime is therefore likely to reduce the probability of an individual engaging in crime (Clarke 1989). Research points out architectural features that increase visibility and accessibility to the library’s spaces, which are suggested to automatically improve visitor and employee safety. But there are other factors: good visual connections and communication between service points (Henrich and Stoddart 2016), good artificial and natural lighting (Henrich and Stoddart 2016), enhanced visibility by rearranging furniture and using transparent/glass walls to create visual connections with closed rooms (Carey 2008; Cromwell et al. 2008; Henrich and Stoddart 2016). In new library projects, it has been recommended that the restrooms should be located close to busy public areas, which would reduce the opportunity for misbehaviour. In old libraries, it “may be the most cost-effective solution” to relocate the library’s staff workstations near to these areas (Cromwell et al. 2008). There are also controversial suggestions, such as having a 24-h café as a strategy to promote surveillance by attracting a “substantial number of patrons at all hours, discouraging acts of vandalism and threats” (Carey 2008).

Beyond changes in the environment, studies suggest using an electronic book detection system at the library exit to avoid book theft (Adewuyi and Adekanye 2011; Cromwell et al. 2008), having guards or library staff circulating often in the libraries (Cromwell et al. 2008; Henrich and Stoddart 2016; Simmons 2018) and using surveillance cameras at key points in the library and external areas (Cromwell et al. 2008; Henrich and Stoddart 2016). Other suggestions to increase safety include using shaming strategies, such as photo displays of anybody caught stealing library materials and body searching of individuals (Adewuyi and Adekanye 2011), especially when patrons leave an area containing valuable collections (Cromwell et al. 2008). For valuable materials, supervised admission and restricted access to rare materials under the close watch of library staff is advocated, supplemented by security cameras to also avoid material damage (Adewuyi and Adekanye 2011; Carey 2008; Cromwell et al. 2008).

Management tactics and staff training are also important measures to prevent crime (Cromwell et al. 2008). Aiming at creating a gun-safe environment, a library in the United States has created a safety committee and the staff has been trained on active-shooter response (Henrich and Stoddart 2016). Similarly, Simmons (2018) mentions the importance of running campaigns urging patrons to never leave their property unattended or encouraging patrons to recruit a library neighbour (i.e. a nearby patron) to watch their belongings until they return. Research also highlights the importance of recording crime and other safety-related incidents in public libraries, so this evidence can later be used to detect potential patterns and help define strategies for prevention (Simmons 2018).

Crime in the library: a conceptual model

Crime and incidents of public disturbance in the library are dependent on multi-scale environmental conditions that are at work at various levels in the building and its immediate surroundings, some of them varying over time. As any other facility, the amount of crime and incidents of public disturbance to which they are exposed depends on the library’s location but also its internal and external features such as good lighting, design and the position of doors, windows and staircases, entrances, types of targets, working routines, management and personal training to deal with visitors and safety emergencies, etc. The internal environment of the library can be divided into six parts that are criminologically relevant:

  1. A.

    Book supply and reading areas These comprise the core of the library, composed of shelves with books and materials. The permeability of the area varies, from restricted areas with rare books to settings with tables for study and reading. The heights and locations of shelves, for instance, will determine the visitor’s capacity for surveillance. Other design and layout features may foster territoriality, for instance, by reducing areas of conflict by controlling access of outsiders (Ekblom 2011a, b; Ekblom and Pease 1995; Newman 1972). Crimes that might occur in this area are thefts of books or visitors’ belongings, such as computers or mobile phones. Because some of these places are hidden places, they provide anonymity for sexual harassment but also robbery. Eating and drinking despite being prohibited in this area might still occur because the visitors are out of sight of librarians and other personnel. Other problems such as public disturbance, sleeping or noise, may be common in certain libraries. Facilities are more vulnerable to crime and acts of public disturbance when they have: many entrances/exits, particularly where they are accessible without passing through the checkout; blind corners, hidden alcoves, high displays that conceal visitors from view and crowded areas around displays of high-risk items (Clarke and Petrossian 2013).

  2. B.

    Transition area These areas may have very different functions, such as corridors, stairs and paths, but can be functional areas, such as computer areas. As in retail shopping, the length and width, location, types of materials, enclosure and design all affect how safe these transition areas are (see for instance, Ekblom (2011a, b). Corridors often have obstacles placed in locations that might offer criminals opportunities to commit a crime and then hide (Newman 1972), e.g. pickpocketing. These areas should be carefully managed as places where drug and alcohol addicts may spend time; they may facilitate drug and alcohol abuse, robbery and problems of public disturbance. Demonstration stalls at corners and in or along corridors can help visitors get to know new books and magazines but also become distracted, and as a consequence, become an easier target for thieves (Ceccato et al. 2018).

  3. C.

    Reception area Contact with the library staff takes place in this area. In many areas, it is where books are picked up or delivered. Personnel are often present and their desks are often placed strategically to promote good surveillance of what happens (Shuman 2002). If not sufficiently trained, personnel can be extra exposed to incidents and conflicts that may occur, which can escalate to serious cases of verbal or physical assault (Kajalo and Lindblom 2016).

  4. D.

    Social area These are the areas where unstructured activities may take place, meaning activities with a variety of noise at particular times. They can be cafés, children’s play areas, storytelling areas, stages for a children’s play or areas where parents park their pushchairs and interact. These are areas where most property and violence crimes are expected to take place, and also where inappropriate use of the area results in acts of public disturbance. Previous research indicates that small distances between tables in a café, for instance, make it easier for thieves to pass, grab a bag and leave (Poyner and Webb (1992).

  5. E.

    Entrance/exit area These areas can be of many types (stairs, corridors, doors), but in any case, well-functioning entrances allow the flow of people both under normal and emergency conditions. Using semi-transparent materials and glass in the construction of entrances allows good sunlight illumination and may affect natural surveillance. The entrances are also the connection with bus stops and lifts. Typical problems are those related to public disturbance, such as people sleeping at the entrance or selling drugs (Ceccato et al. 2022).

  6. F.

    Surrounding areas These areas are visible from the library in a field of view between 50–100 m (half a block to 1 block) from the entrance(s) of the building (note that around 50 m is the distance at which a person can still recognise others and also clearly hear shouts for help, see Gehl (1987). High-crime areas tend to affect victimisation at libraries and perhaps vice-versa, as Bowers (2014) suggests, there is a positive relationship between internal and external crime in an area. Libraries that are located in central urban areas may be extra vulnerable to crime spill-over from mixed land use, e.g. with criminogenically and badly managed bars and restaurants (Ceccato et al. 2018; Eck 2019; Newton 2018). Therefore, librarians together with other place managers (private and public) play an important role in keeping records about crime and incidents of public disturbance in this common public area. The municipality, and urban planners in particular, together with security experts, can cooperate with neighbourhood leaders and other place managers to deal with long-term safety problems of the areas surrounding libraries.

We adopt here the proposed conceptual model conveying the idea that the design and efficient use of the library environment by visitors and staff are essential when it comes to promoting safety conditions, both by preventing crime and public disturbance incidents and by affecting how the library is perceived by personnel and visitors. The three research questions of the study are:

  1. 1.

    Which are the most common types of crime and incidents of public disturbance in libraries?

  2. 2.

    Do incidents of crime and public disturbance in libraries vary spatially and temporally? Are there differences between libraries? Based on evidence from a three-dimensional model and fieldwork inspection, which are the library environments most in need of crime prevention intervention?

  3. 3.

    Which are the experiences and perceptions of safety problems declared by the libraries’ personnel? Do they feel safe at the library?

Study area

Stockholm has about 40 libraries (depending on how libraries are counted), distributed among its 13 districts. Two libraries were chosen as study areas: “Stadsbiblioteket” (which translates to “The City Library”) which is the largest library in Stockholm (with around 800,00 visitors annually) and located in a busy central area, close to a main road, metro stations, a university, bars and restaurants, and “Högdalen” (around 190,000 visitors annually) which is a ‘metro-library’, as it is connected to a metro-station in the outskirts of Stockholm. Both libraries have a similar number of days when they are open, around 340 days per year and are surrounded by relatively criminogenic environments. Security guards may be present when they are called by personnel. Beyond location and size, these libraries also differ in layout and design, and in the activities that they promote. Note that these libraries may not represent all libraries in the city but are good examples of two extremes: one library is large, located in the city centre and the other is relatively small, located in the outskirts of the city, and linked to a metro station.

The first library, Stadsbiblioteket, is about 3000 m2 (not all public areas) with two floors open to the public and two entrances (one in the middle of the main façade of the building). The first floor (entrance floor) houses the children’s area, a coffee shop, a reception, restrooms, lifts and stairs that lead to the second floor. The second floor has a central circular area, called a ‘rotunda’, where it is possible to find an information desk, books and a stage for cultural events. The location of the information desk in the rotunda functions as a ‘panopticon’ where personnel have a central overview of what happens in the library. The second floor also has restrooms, reading areas, a copying room and lifts and stairs that connect to the first floor. The computer and study areas are distributed in different areas of the library. The second library, Högdalen, is smaller with around 600 m2 of public space. The design of the library was changed in 2019 to an open plan, but with some areas designated more clearly for specific activities (e.g. for reading, or for children). There is only one reception where the personnel assist the public and it is currently located near the computer and printing areas, under personnel supervision. The two entrances are linked either directly to the metro or the street or connected to it more indirectly, where a bus station is located. The connection with the metro-station makes the library accessible from inside the station and from the street. The library is located close to a busy bus station, cafés, restaurants, supermarkets, an alcohol sales outlet (Systembolaget) and close to a car park.

Data and methods

Data

Different datasets were used in this study.

  1. 1.

    Incidents data was collected by the personnel of the two libraries between 2017 and 2020 and contains two sets of information: one is composed of the date, time, and nature of the incident (crime and other safety-related incidents), and the second was a column with a detailed description of what happens in each event written as a free text.

  2. 2.

    The estimated number of visitors during opening hours for the two libraries centrally recorded provided by Stockholm libraries from 2017 to 2020.

  3. 3.

    Blueprints with floor plans of the two libraries.

  4. 4.

    Data collected through fieldwork inspection performed in a series of visits to the libraries (July 2021–May 2022), focusing on the inspection of particular environments most targeted by crime and incidents of public disturbance.

  5. 5.

    Eight interviews with the personnel of each library were carried out between May and October 2021 (Five at Högdalen and three at Stadsbiblioteket) using a convenience sample. They were all permanently employed (as opposed to temporarily).

Methods

Preparation of incidents database

The dataset contains a column with free text with a description of all incidents. The data used was registered by personnel in the libraries. It is unclear whether reporting practices by the personnel have affected the datasets but researchers learned that a course in 2018 provided guidelines to the library personnel about how to record crime and safety incidents in the system. New library personnel learned the recording system from current library staff.

Initially, we noticed that there were many multiple incidents recorded as one. Therefore, we split these double/triple incidents and re-coded them. For instance, first an act of public disturbance (visitor was shouting in the children’s area at 17:32), then he moved to another room causing vandalism (damaging newspapers in a book area at 17:46), and finally, the visitor used abusive language towards the staff of the library (verbal aggression at reception, at 17:55). From one incident, we divided the record into three incidents, sometimes with same/similar or different time and place stamps, depending on the information available. Table 1 and Figs. 1, 2 and 6 show examples of the types of records. The database was re-classified by three researchers independently, using a cross-validation of the records, with a 78.6% matching rate of total incidents (this means that 21.4% of the records were recoded based on a discussion among the three researchers to decide to which category a particular incident should belong until 100% agreement was achieved). This procedure was carried out to make sure the recoding was the same independently of the researcher. The dataset was re-classified into four main categories and several subcategories: (1) Public disturbance, e.g. inappropriate use of space, sleeping in the library, noise, smell; (2) Vandalism, e.g. damaging the building or objects in the library, including books; (3) Aggression, e.g. any kind of physical or verbal aggression against employees or other visitors, and (4) property crime, e.g. theft, robbery or pickpocketing. A fourth of all incidents was excluded from the analysis because they were neither a crime nor events of public disturbance (e.g., malfunctioning lifts). Note that the database may suffer from biases because it is dependent on a staff decision to record incidents and therefore staff-to-staff incidents may not be present in the data.

Table 1 Examples of crime and incidents of public disturbance in Högdalen and Stadsbiblioteket
Full size table

Visualisation in CAD

Blueprints with floor plans were used to create three-dimensional models of the buildings using CAD (Computer-Aided Design). Each floor was considered a separate layer connected by limited-access corridors (or lifts and stairs) that could also be visualised together and used as the basis for analysis. Then, the model was populated with crime data using percentages for both libraries. Of the total incidents, about 10% had no location and were not mapped. Places that appear more than once in the map were mapped with the same value, as the information available does not specify, for example, in which particular restroom or lift the incident occurred. Incidents that happen at an “entrance” (i.e. no particular entrance was specified) were mapped proportionally (two-thirds were associated with the main entrance and one-third with the secondary entrance). The analysis of location and frequency was carried out using spreadsheets.

Fieldwork inspection

The libraries were inspected on several occasions by the researchers from July 2021 to May 2022. Using the conceptual model as a reference (Fig. 1), a field work protocol was utilised to inspect the libraries (Ceccato 2019). Natural surveillance was assessed based on the visitors’ ability to see what happens in a particular area of the library (sightlines), which was highly affected by the design of the setting, and environmental and temporal contexts. How visitors use spaces to communicate ownership in the library was an indication of territoriality. Access control was evaluated by the existence of entrance checks (mechanical), security guards or personnel at the entrance. Means to increase the efforts that offenders must expend in the commission of a crime was an indicator of target hardening while maintenance routines were an indicator of library image. Finally, the use of design and signage to encourage intended patterns of usage of public space was investigated in activity support of the library.

Fig. 1
figure 1

Crime and safety incidents in libraries: The A–F library conceptual model

Full size image
Fig. 2
figure 2

Crime and public disturbances in Högdalen (N = 296) and Stadsbiblioteket (N = 1178), 2017–2020

Full size image

Interviews

Interviews were conducted with permanent, well-experienced library staff having worked at least three years in these facilities (five in Högdalen and three in Stadsbiblioteket), who were predominantly women (five women and three men). The semi-structured interviews utilised the A-F library model (Fig. 1) as a reference and included open-ended questions in order to give the respondents a chance to formulate their own answers and perceptions on the following subjects:

  1. (a)

    the personnel’s personal experiences of the library’s working environment;

  2. (b)

    the personnel’s personal experiences of safety problems encountered at the library, focusing on the library environment and their training;

  3. (c)

    the personnel’s perceptions of safety; and,

  4. (d)

    the working methods and routines at the libraries to support the personnel’s work in dealing with safety problems.

Due to the world situation with Covid-19 at the time, all interviews were conducted digitally. Recording and transcription of the interviews were used in the study to achieve authenticity.

Results

Nature and frequency of crime

In total, there were 296 incidents recorded in Högdalen library and 1178 incidents in Stadsbiblioteket library from 2017 to 2020. Although events of public disturbance dominate the records in both libraries (56%), violence and theft/robbery are more common in the central city library (Stadsbiblioteket, 0.8 cases per day, 60 per cent of days with no incidents) than in the metro library (Högdalen, 0.2 cases per day, 84 per cent of the days with no incidents).

In Högdalen, there were 75% of acts of public disturbance, 7% of vandalism, 1% of theft and 17% of aggression, while in Stadsbiblioteket, there were 50% of acts of public disturbance, 5% of vandalism, where 18% were thefts and 27% aggression. Within each crime category, there are variations in event type, such as attempted shoplifting and burglary but also use of weapons, and explosives. Violence and aggression involve a range of offences from verbal to physical attacks and threats against personnel to sexual harassment. Table 1 illustrates the common incidents recorded by the library personnel, see also Fig. 2.

Temporal variation of crime

Figure 3a shows that most crimes and public disturbances occur during the afternoon. The number of events is generally greater in the middle of the week and less frequent at the weekend (Fig. 3b). They are also more frequent in the winter months such as January and February (Fig. 3c), even after being weighted according to the number of visitors.

Fig. 3
figure 3

Temporal variation of events in Högdalen and Stadsbiblioteket 2017–2020. a Number of events per opening hours; b Percentage of events per days of the week; c Events per months of year. In colour the percentage of events per month/year, dashed line indicates the rate of events per 1000 visitors

Full size image

Most events of public disturbance in the libraries occur in the afternoons, but start increasing around 11 a.m. There is a peak in mid-afternoon (3–6 p.m.) in Stadsbiblioteket and at 3 p.m. in Högdalen. Public disturbance is evenly distributed across weekdays and over the year. For property crimes, most thefts occur between 11 a.m. and 5 p.m. with a peak at 3 p.m. Tuesday is the weekday with the greatest number of this specific event as well as February with a number almost 50% greater than other months. (Thefts and robbery were not reported in Fig. 3 because the numbers were too small). Vandalism occurs mainly in the afternoon and more frequently between 4 and 6 p.m. Thursday is the day with more vandalism events registered in both libraries. Högdalen does not have any vandalism reported on Saturdays, while Stadsbiblioteket does not have any vandalism recorded on Sundays. In Högdalen, vandalism occurs mostly in the first 6 months of the year (January, March, April and May), while in Stadsbiblioteket these events peak in February and October. Acts of aggression are evenly distributed across the hours of the day (with a slight peak at noon in Stadsbiblioteket and around 3 p.m. in Högdalen) and weekdays, with the highest number of events on Wednesdays and Thursdays in both libraries. Conflicts, aggression and other acts of violence acts are fairly evenly distributed over the year in Stadsbiblioteket, but with relatively high numbers in February and October. In Högdalen, more than 60% of these types of incidents occurred in January and February (Fig. 3).

Spatial variation of crime

The geography of crime and public disturbance incidents varies in these two libraries, as is illustrated in Fig. 4. The shades indicate the distribution of the incidents by percentage (as a reminder, around 10% of events were not mapped due to a lack of information). In Stadsbiblioteket, the place with highest percentage of incidents (56%) is the Book Supply and Reading Area, while in Högdalen it is the Entrance/Exit with 27% (Figs. 4 and 5). In both libraries, the Transition Area has around 20% of the total number of incidents followed by the Reception and Surrounding Area where the smallest percentages of crimes were reported. This can be partially explained by the fact that personnel do not frequently record incidents that occur outside the library’s door and near to the reception, where personnel work, as fewer incidents occur, apart from conflicts with visitors.

Fig. 4
figure 4

Spatial variation of crime in Högdalen and Stadsbiblioteket 2017–2020 based on the A-F library conceptual model

Full size image
Fig. 5
figure 5

Spatial variation of crime in Högdalen and Stadsbiblioteket 2017–2020. Percentage of incidents per section. Note that in Högdalen, more than 20% of incidents were concentrated in particular sections of the library, in the book supply/reading area (including children’s area)

Full size image

The Social Area is composed of settings that represent the greatest difference in frequency of crime between the two libraries, 2% in Stadsbiblioteket but over 10 times higher in Högdalen. (Note that the Social Area in Högdalen also represents around 10% of the library’s total area, while in Stadsbiblioteket it represents only 4%). In both libraries, public disturbance is the most frequent type of incident across all settings, except for the Reception in Stadsbiblioteket, where aggression is the most common type of crime (Fig. 6a). In Högdalen, aggression is greater outside the library premises, namely in the Surrounding Area of the library (Fig. 6b). Vandalism seems to be more homogenously distributed in Högdalen, except for in the Social Area and Surrounding Area, where there are no records. The opposite occurs in Stadsbiblioteket, as vandalism incidents are highly concentrated in the Surrounding Area of the library, perhaps reflecting the central location, surrounded by restaurants, bars, a university and two metro stations. In Stadsbiblioteket, crimes against private property most frequently occur in the Social Area (20%) and in the Book Supply and Reading Area (20%), while in Högdalen the number of incidents were very few. Figure 6 illustrates in detail the distribution of incidents by library area.

Fig. 6
figure 6figure 6

a Counts and percentages of crime and incidents of public disturbance based on the A-F library conceptual model, Högdalen, 2017–2020. b Counts and percentages of crime and incidents of public disturbance based on the A-F library conceptual model, Stadsbiblioteket, 2017–2020

Full size image

Crime and safety at the library: the personnel’s perspective

The types of problems revealed in interviews with the staff are similar in both libraries but there is a recognition that aggression and other types of violence occur more often in Stadsbiblioteket than in Högdalen, where events of public disturbance are also common. Visitors who consume or have consumed alcohol and/or drugs are often mentioned as a problem in both libraries. The types of events vary as described below and can directly affect the staff’s safety.

“There can be conflicts between visitors and staff.” (H1)

“We have a lot of trouble with messy young people and also alcoholics.” (H5)

In both libraries, the staff say that they have to deal with issues of safety in the library at least once a week. More than one interviewed person stated that “these situations come in waves” and that they can mirror what is happening in the community and/or surrounding areas, such as problems in the local schools, which directly interfere with the library’s routine.

“… it may not be every week, but once or a couple of times a month.” (H3)

“You often have the experience that sometimes everything happens at the same time and then there can be long periods where it is very calm.” (H1)

The staff declare that they can distinguish specific behaviour that makes it necessary for them to take precautionary actions, at particular times and/or in particular places. Conflicts increase around the library’s closing time because visitors may not want to leave the premises:

“There is a sensitive moment during the day, and it is when it is time to close [the library].” (H3)

“when people are under the influence of drugs or under the influence of alcohol” (H3)

“when there are large groups of young people” (H4)

“situations when there is conflict between visitors” (H4)

Most of the staff declare feeling safe in the library, but mention some situations which make them feel uncomfortable and are hard to deal with. They recognise that these situations are part of ‘dealing with public places’ and that these situations can be managed, but they do not always feel prepared to do so. They mention cases in which groups of visitors under the influence of alcohol and/or drugs challenge them, which make them feel unsafe, as illustrated by:

“people with alcohol and drug problems create a feeling that anything can happen because that person can act unpredictably.” (H2)

“I felt unsafe several times. Most cases are linked to young people staying here when it is cold outside.” (H1)

Those few members of staff who declare feeling unsafe at work also declare feeling powerless, mainly for being women and feeling intimidated by groups of young men. It was also mentioned that the prejudice towards young male immigrants who harassed personnel has increased in the last decade. The most extreme case was one member of staff who felt a higher level of stress related to the conflicts in the library.

“I feel angry, frustrated that I worry about conflicts at the library when I am at home and should have my own time.” (H1)

“You feel quite powerless when you are harassed in that way as a woman and that makes me feel prejudice against immigrant boys.” (H5)

The staff indicate the importance of the physical and social environments for safety conditions in the libraries by creating activity nodes and stimulating appropriate behaviour. The centrally located position of the reception in Stadsbiblioteket and the zoning in Högdalen are two examples, although this is clearer in Högdalen than in Stadsbiblioteket. In Högdalen, the recent improvements in natural surveillance (in particular, lower bookshelves and improved placement of the information desk) and increased territoriality through zoning (clearly demarcating the children’s area) were mentioned by staff:

“We have worked a lot with zoning, where we then, for example, moved all the computers to the library entrance in an area far from the area designated for adult literature where people want to sit and read…. but the zones are not fully adapted for adults…there is no good place where you can gather for a book circle.” (H3)

The zoning is intended to create micro-places where particular activities or behaviours are encouraged through the adequate use of the design of the library, such as the position and the height of the shelves. The zoning can be useful to put the libraries’ multifunctional role to work but still may not be enough to reach the expectations and needs of all visitors, as illustrated by the staff quotes above.

The staff feel prepared to act if anything occurs, which may involve taking precautionary measures, such as closing the library at particular times or always working in groups:

“We never work alone out in the library. And when we close, there are always three of us, at least three, precisely because things can happen at the end of the day and then you should not stand there alone but then you must have some colleagues that you can lean on, who can be helped.” (H5)

“There were a few times when we closed the library earlier so that the young people would miss their time when they usually come and mess around.” (H1)

The goal is to approach a situation in a polite and friendly way, to try to give visitors “options to solve the issue without increasing the problem” (according to the staff). However, if the situation gets out of control, the personnel contact the guards, who, according to the interviews, normally get to the library quickly (although at Stadsbiblioteket, the police may be also called). It was also mentioned that staff training also involves rare events such as violent attacks or bombs. Challenges mentioned regarding staff training included frequent personnel changes (i.e. the need to train new staff) and that the current Swedish training of librarians neither involves any conflict management training nor considers the demands of a library as a multifunctional public place. The interviewed staff also feel that increasing cooperation between the library and other local stakeholders is an important measure to ensure a safe library.

Discussion of the results

The overwhelming majority of incident records in Högdalen and Stadsbiblioteket are minor incidents of public disturbance. Aggression (conflicts between visitors, or between visitors and personnel), property crimes (such as thefts and robberies) and vandalism are less commonly reported. Similar pattens can be found in Sweden for other types of facilities, such as stations, parks and shopping centres (Ceccato & Uittenbogaard, 2014; Iqbal & Ceccato, 2016, Ceccato et al., 2018). Most crimes and public disturbances peak during the afternoon hours (around 3 p.m., after schools close), on Wednesdays and Thursdays and during winter and spring, reflecting people’s routine activity (Cohen and Felson 1979). These daily variations reflect the inflow of visitors, particularly of students after school hours when youth and/or associations are closed. In addition, winter and spring days attract more visitors than the other seasons in a country with long dark winter days and when schools close. The library gets an overflow of visitors during these short breaks. Locational factors also affect the types and levels of safety incidents in libraries as it does the library’s size and number of floors. Being larger and centrally located, Stadsbibliotektet reported more incidents per visitor than Högdalen between 2017 and 2020. As expected initially, libraries that are located in central urban areas may be extra vulnerable to crime spill-over from mixed land use, e.g. bars and restaurants.

The design and layout of libraries play a role in their criminogenic conditions. Some events occur in hidden places, such as the restrooms in Stadsbiblioteket, that could be considered enclosed rooms. The crimes in these areas represent around 10% of the total of crimes in the building, even when they represent less than 1% of the library’s total area. Incidents also occur in the stairwells in the Entrance/Exit of Högdalen library, where is a difficult area to control as it is outside the staff’s sightline. Similarly, in the literature, Henrich and Stoddart (2016) mention stairwells as potential locations for crime; entrapped and enclosed rooms were also suggested as potential target settings (Carey 2008; Henrich and Stoddart 2016).

Areas with poor natural surveillance, such as behind and between bookshelves, have great potential for crimes (Cromwell et al. 2008). In a single spot in Högdalen library (corner Book Supply area), 8.8% of all the incidents registered in the building were concentrated, and its location is isolated from the reception desk. The same applies in the hallways in Stadsbiblioteket which are also used as Reading and Book Supply areas and to which the personnel do not have continuous access (Carey 2008; Cromwell et al. 2008; Henrich and Stoddart 2016). Corridors, dark areas and corners that lack proper illumination are also areas of concern and constitute what is classified as a Transition Area. In both libraries, rooms with valuable collections or “CRAVED” products such as computers are also vulnerable places for thefts (Carey 2008; Cromwell et al. 2008; Smith 2018), but less so in Högdalen where very few cases were reported.

Some places are criminogenic because of the activities they attract. For example, in Högdalen, the Entrance/Exit attracts a high number of incidents (a fifth of the total), which is where people spend time on stairs, sleeping or drinking, waiting for buses, or are on the way to the metro station entrance. Previous studies have shown that facilities are more vulnerable to crime when they have many entrances/exits, particularly where they are accessible without passing through the checkout (Clarke and Petrossian 2013). Seasonal, weekly and hourly variations in flows of people in the surrounding areas (metro, schools, etc.) affect what happens inside the libraries, creating ‘waves of safety problems’. Areas of increased safety concerns in both the studied libraries are the children’s areas, where the most diverse clientele can be found (see also in the literature, Carey 2008). Such areas attract high numbers of public disturbances, which is also the most common type of incident recorded in both facilities. In Stadsbiblioteket, the children’s area and the rotunda (central space on the second floor, with diverse uses such as Book Supply and Reading and Social Area), a number of incidents are concentrated, including violent encounters and thefts. In both libraries, there is little specification of what property was taken; however, these events occur mainly where people tend to use their personal computers and stay for longer periods, such as in the Reading Area and Social Area of Stadsbiblioteket (each with 20% of the total of Property Crime). Similarly, Simmons (2018) suggests that criminal activities related to patron’s property theft occur mainly when they leave their belongings unattended (despite signs warning them of the risk of theft), which can often occur in these particular areas of the library (as also confirmed in the interviews with the libraries’ staff).

Responses from the interviews indicate the role of library staff as place managers (see for instance, Linning and Eck 2021) and therefore their active role over what happens in these facilities and, to some extent, in the surrounding areas. Overall, most of the staff declare feeling safe in the library but indicate situations in which they felt threatened. Dealing with people under the influence of drugs and/or alcohol, or being a woman and feeling intimidated by groups of young men are examples of situations in which they feel powerless and unsafe. Personnel are well aware of the temporal variation of these safety problems (some  of these problems come “in waves”) and take precautionary measures to avoid them. They indicate the importance of the physical and social environments for the safety conditions in the libraries, especially in Högdalen where the environment (improvement of natural surveillance, territoriality) and staff training (activity support) were discussed by the interviewees. The staff calls for the need for adequate training as part of formal librarianship education to learn how to deal with the current safety demands in the library setting.

Conclusions

A conceptual model has been proposed based on environmental criminology theory and principles of Crime Prevention Through Environmental Design (CPTED) to illustrate the nature of crime and incidents of public disturbance in libraries in Stockholm, the capital of Sweden. The methodology applied makes use of three-dimensional visualisation using CAD to detect places where safety incidents are concentrated in micro-places. With these diagnostics in hand, a fieldwork inspection was then performed to investigate the characteristics of the most problematic areas (as shown in Fig. 1), combined with the library staff’s perspectives on the safety conditions in the libraries.

Findings show that incidents of public disturbance by far constitute the most frequent safety problems in these libraries followed by aggression, property crime, and vandalism. Any safety interventions to deal with these incidents should consider the heterogenous nature of these incidents—some are criminal but the vast majority are not. At the same time as librarians embrace the new role of libraries as a multifunctional meeting place, they also face pressure to maintain the core activities of the library, lending books and being a place of quiet and reading. Someone speaking loudly on the streets would rarely attract any attention from security guards, but in a library such an act disturbs visitors. Thus, the current multifunctional role of libraries imposes new challenges on those who work at these establishments.

These incidents are not distributed randomly in space or time (hourly, weekly, seasonally). Maps and figures show that most problems occur in relatively few micro-places, but vary by incident type. There is also some evidence of specialisation of incident location, such as conflicts occurring in the reading area/children’s area and public disturbances in entrance areas. The three-dimensional visualisation helps illustrate how crime is concentrated in space, which corroborates previous research on crime concentration in micro-places. However, such concentration (in the children’s area, book reading area and entrances) appears at lower levels than those found in e.g. stores in shopping centres (Ceccato et al., 2018) or in street segments (Weisburd 2015). This is certainly because, compared with stores in shopping centres or thousands of street segments in a neighbourhood, micro-places in libraries comprise limited and enclosed environments that are often few in type. Findings also illustrate the potentiality of using CAD in environmental criminology in particular when multi-story buildings constitute the study area. Future studies could further explore the time dimension using new generations of CAD, possibly using virtual reality in experimental frameworks to test improvements of the environment.

The design and layout of libraries play a role in the libraries’ safety conditions (e.g. the presence of territoriality to avoid conflicts and clear sightlines above and in between shelves promoting surveillance opportunities for personnel), but some places are more challenged than others because of their poor design (e.g. open computer areas, location of cafés) or because of the activities they attract (e.g. the type and number of entrances are permeable spaces that can be affected by activities from the library’s surrounding areas). The interviewed staff declare feeling safe in the library but mention unsafe situations (places/times/people) which they avoid by taking precautionary measures and/or training courses to manage these safety challenges.

With the help of the fieldwork inspection, it was possible to obtain clues about why similar places ended up having different levels of crime and/or incidents of public disturbance. A list of micro-places most in need of intervention is suggested using the conceptual framework proposed in this study, which is discussed in detail in the next section. Although the inspection of the library is not conclusive as to how the environment ‘promotes’ crime and incidents of public disturbance, it provides detailed clues as to how particular micro-places can be more vulnerable to crime and incidents of public disturbance than others. This limitation calls for caution when drawing conclusions about the causality between the occurrence of crime and particular micro-places in libraries. Future studies could test the A-F conceptual model in other types of libraries.

Moreover, as in any other mapping methodology, the detailed analysis of the public library is highly dependent on accurate, precise and complete datasets. Having a systematic recording system of the incidents by personnel over time at micro-places in the library is therefore fundamental. In this study, some re-coding of incidents was necessary to create the locational information of these events from a general database (free-text description of incidents). Around 10% of records were not mapped because location information was not available. For example, the incident was described as happening ‘in a restroom in the library’, but the record did not specify which restroom. Similar problems are encountered for events with missing time stamps. Another limitation in this research is the inaccessibility to researchers of ‘staff-only areas’ in the library. Data permitting, future research should devote time to testing, in a more controlled manner, the potential effect of public libraries’ micro-places on crime and safety (of visitors and staff), using more libraries of similar sizes and geographical contexts. Finally, relatively minor attention has been paid to the need to distinguish between empirical research, on one hand, and practice-based checklists which have not been tested empirically on the other.

Several other related topics would be important to explore in future studies. It would be helpful to estimate a frequency of safety incidents per visitors, or per staff member to get an idea of the extent of possible victimisation of staff. The potential harm of crime or events of public disturbance such as financial costs or emotional impacts on staff of public libraries still remain an area for further research. In addition, future research should include correlation analyses between the frequency of crime in those libraries and the frequency of crime and events of public disturbance observed in the post code districts where they are located. There is also a question about potential differences between crime patterns in public libraries as opposed to university libraries, which has not yet been studied. Finally, future research should split crime and events of public disturbance. There is also a need to further problematise the issue of incidents of public disturbance. There is extensive literature on policing public spaces which shows that policing which concentrates on acts of public disturbance rather than actual crimes can be discriminatory (against young people and/or minorities), as a reference, see for example the work by Wakefield (2004) on cultural centre, shopping centre and retail and leisure complex. The effects of this type of governance need to be further investigated as they are being diffused further locally, due to the growing collaboration between private security, the police and civil society.

This study has a few important theoretical contributions to routine activity and other situational approaches. First, we showed that libraries work as a crime attractor/generator in an area but they may also function as a crime preventive motor. The library attracts young people that have nowhere to go after school, therefore, libraries unintentionally support initiatives of crime control. For CPTED principles, there are a few impacts. Although not all principles of CPTED are equally important to support the understanding of safety conditions inside the libraries, the principles are important starting theoretical markers (e.g., clear links between natural surveillance and the location of shelves and reception desks). Yet, if situational principles continue to be applied to libraries, then more attention should be given to the role of urban contexts in these facilities, as our findings show evidence of the links between indoor conditions and neighbourhood contexts.

Implications for practice

The library environment should follow basic principles of situational crime prevention (or more specifically, CPTED principles) from the start. Most CPTED techniques and personnel training can be introduced without large investments. Our findings show that it ‘pays off’ to plan for libraries to have a good overview, where the premises are designed with clear zoning and signage for increased orientability. There is also a need to proactively plan efforts to reduce risks in places with conflicting goals. For example, a quieter zone should not be placed next to a busier one, like a children’s area or cafe. Libraries, like all other facilities, follow a rhythm which means that crime and disturbance can vary in time, per hour of the day, day of the week, month and/or season, which requires knowledge of different challenges but also training and flexibility of staff. It is important to evaluate (continuously and over time) the safety of libraries’ internal environments in relation to changes in the profile of visitors by e.g. age, gender, gender status and disability.

As our examples show, public library entrances and surrounding areas can be particularly problematic. Visibility from indoors is a key safety factor. Using semi-transparent materials and glass in the entrances allows good sunlight illumination, improving natural surveillance by staff and visitors. The entrances are also the connection with e.g. bus stops and lifts, where typical problems are those related to social disturbances, such as people sleeping at the entrance or selling drugs near the library. This surrounding area constitutes a grey zone in which the responsibility for safety in these environments is diluted between place managers and the municipality. The police, security guards and library staff all have an important role to play in keeping records about incidents in this common public area. Cooperation among neighbourhood leaders, transport operators and other key stakeholders is important when it comes to dealing with long-term safety problems in the surrounding areas of libraries.

This study is unique because it makes use of recorded data to create a good knowledge base about actual safety problems before starting crime prevention work. We recommend public libraries in other parts of the world to ensure that any safety incident is systematically reported in a common system shared with other libraries. This demands the development of clear routines and follow-up of incident reports, as well as a simple and accessible system to report safety related incidents. These records can then be regularly discussed at work environment meetings with safety representatives. The use of safety surveys with visitors and interviews with personnel can also support the work to ensure a safe and secure public library. Finally, the staff in our interviews point out how important it is to have the right training to be able to handle and de-escalate difficult situations themselves before calling in security personnel. In the long term, librarianship education could better prepare students for service, hospitality and safety to a greater extent than is currently the case.