Compassionate revanchism: The blurry geography of homelessness in the USA

Introduction

In this paper, we argue that spaces of care (shelters, charities, etc.) in revitalising cities often bolster anti-homeless urban geographies despite their stated aim of promoting compassion. Similarly, we found that some punitive institutions (i.e. police) adopt more caring means of control. Based on studies of Fresno, California, and Phoenix, Arizona, we show how spaces of care remained vulnerable to the geographic pressures of urban capital, such that service providers and police departments mutually worked towards a common goal of managing homeless people’s movements in the city. Such mutuality, we argue, regularly blurs any division between care and criminalisation. Through situating the mutual development of shelters and anti-homeless policing in greater Phoenix, we highlight how care and criminalisation meld amidst the broader motives of revanchism. In Fresno, we show how urban revitalisation pressured homeless shelters to champion the city’s punitive policies, such that care workers often unwittingly and indirectly became architects of revanchism. Together, these examples highlight how so-called spaces of care deserve greater scrutiny in revitalising cities. They also force us to consider how such spaces may function – as with anti-homeless laws – to service revitalisation.

Until recently, much of the literature on US homelessness argued that the intensifying commodification of urban space creates a punitive landscape of homeless criminalisation and exclusion (Davis, 1990; Mitchell, 1997; Smith, 1996). Yet urban scholarship now overwhelmingly resists the punitive framework and characterises homelessness management as a complex, nuanced, and often accommodating system of institutions (Cloke et al., 2007a; Conradson, 2003; Deverteuil et al., 2009; May and Cloke, 2013; Murphy, 2009; Parr, 2000). These authors urge scholars to examine the myriad spaces of care in the city, spaces where ‘proactive interest of one person in the well-being of another’ occurs and is expressed by both the physical (clothing, feeding, healing) and emotional labour of unconditional support (listening, offering advice, withholding judgement) (Conradson, 2003: 508, 512). Building on both ‘camps’ within urban studies, we show how spaces of care are themselves shaped by the punitive tendencies of capitalist urbanisation. In this light, analyses of care through individual or institutional sentiment can lose sight of how spaces of care are vulnerable to economic pressures to remove homeless people from public space. As Smith (1996) argues, revanchism is not simply personal animus, but a structural tendency of urban revitalisation to displace and exclude poor and homeless people. Further, though Deverteuil et al. (2009) acknowledge that the ‘abeyance’ or warehousing function of shelters coincides with anti-homeless policing, the empirical dynamics of abeyance have remained largely unexamined. We intervene in this literature to show how ostensibly compassionate agencies in revitalising cities often display a structural tendency to control and exclude unsheltered homeless people.

Critiques of the revanchism thesis not only emphasise the importance of care, but also highlight the problematic tendency of relying on studies of New York City and Los Angeles to make broad claims about the nature of urban space (DeVerteuil et al., 2009; Johnsen and Fitzpatrick, 2010). Further, such critiques contend that while US revanchist politics followed a period of disinvestment in homeless services, it is important to examine the urban dynamics of cities that have robust social services. In responding to these critiques, we turn our attention to two cities that have received very little attention in the literature on homelessness or urban studies, and that have experienced vastly increased service provision over the past decade. Drawing also on Stuart’s (2014) emphasis on the importance of examining homelessness in marginal urban spaces, we contend that it is equally important to examine marginal cities themselves. Fresno and Phoenix offer helpful case studies in that they are understudied cities that have also been lauded as models of care and service provision. Indeed, Phoenix’s homeless campus has been praised as a ‘national model’ (Hermann, 2005) and federal officials called Fresno’s homelessness management ‘a shining example of what’s possible’ (Appleton, 2015). Finally, we take up DeVerteuil’s (2014) call for more complex, fine-grained ethnographies of homelessness management across locations. The analysis of Phoenix draws upon ethnographic work as a case manager in a drop-in centre over 2009–2010 and ten interviews with service providers and homeless individuals in 2012. The analysis of Fresno draws on three months of ethnography conducted in 2013, as well as 24 interviews with shelter operators, public officials, activists, and homeless Fresnans. Both case studies involved exhaustive archival research of government documents and local newspaper reports up until 2013. Through interrogating abeyance in these two cities, we argue that compassionate motivations and explanations do not always produce favourable outcomes for homeless people. And this tendency emerges, we stress, because homelessness charities and police departments remain, together, pressured by the needs of urban revitalisation.

Care and punishment in the urban governance of homelessness

Scholars have traced a marked shift in the US urban landscape since the twilight of Fordism and dawn of neoliberalism, revealing an urgent set of policy priorities designed to attract new forms of economic investment. This process, Harvey (1989) argued, produced ‘entrepreneurial cities’, municipalities increasingly focused on selling the landscape as ‘business friendly’ to stoke redevelopment (Hall and Hubbard, 1996) rather than providing public services (Cox and Mair, 1988; Logan and Molotch, 2010). Mayer (2012) has argued that this deepening of urban commodification has only spread – however unevenly – as cities engage in inter-urban competition for footloose capital. Endeavouring to advertise an improved, up-and-coming image, entrepreneurial cities use a range of practices to reduce the presence of marginalised groups (MacLeod, 2002).

As such, laws were drafted that targeted the activities that homeless individuals undertake to simply survive: panhandling, sitting on sidewalks, ‘camping’ in public spaces, and pushing shopping carts (Amster, 2008; Mitchell, 1997; Smith 1996). In some cases, homeless people were ‘banished’ altogether by trespass laws (Beckett and Herbert, 2010), often only able to find respite in jails, shelters, drop-in centres, and welfare offices (Stuart, 2014). Hitting their stride in the 1990s, such anti-homeless laws were matched with funding for their enforcement. Aggressive arrest campaigns, property sweeps, and the bulldozing of encampments ascended, becoming normalised police practice. The criminalisation of homelessness in the USA has only continued to rise in the last decade, with a 43% increase in laws against sitting and lying and a 60% increase in city-wide bans of homeless behaviours since 2009 (National Law Center on Homelessness and Poverty (NLCHP), 2014). As Davis (1990) notes, anti-homeless laws are matched with anti-homeless architecture, embodied in ridged planters, gated communities, and ‘bum-proof’ benches. Specifically designed for – or particularly enforced within – the centres of redevelopment, these strategies safeguard the economic growth of urban spaces of consumption (Mitchell, 1997).

Accordingly, and overwhelmingly drawing from the US experience, a significant amount of homelessness research has focused on these punitive trends, illuminating how the criminalisation of homelessness produces a shrunken, interstitial, isolating, and unstable geography for homeless people. Yet this punitive framing, DeVerteuil and colleagues (2009) argue, focuses too squarely upon the ‘collapsing’ geography of homeless people. ‘Far from any collapse,’ DeVerteuil et al. (2009: 652) explain, ‘there has in fact been a remarkable proliferation of homeless spaces over recent years, including the considerable expansion of abeyance, ambiguous and interstitial structures that accommodate and support homeless people’. As important as studies of the punitive framing are, DeVerteuil et al. argue, homelessness research must include ‘complexity’, looking not only at the eradication of places of homelessness but their transformation, and how the punitive turn is accompanied by the proliferation of ‘spaces of care’. As Cloke et al. (2007b: 390) argue, spaces of care emerge ‘in the interstices of revanchist space to provide comfort and care to the excluded, including the homeless’. It is for this reason that scholars have demanded a greater attention to the myriad geographies of care and inclusion in cities, such as homeless shelters and treatment facilities, rather than only criminalisation and exclusion (Conradson, 2003; DeVerteuil, 2006; Johnsen et al., 2005; Laurenson and Collins, 2007).

Homeless shelters and other sorts of services, then, have been and continue to be of interest, but have also been subject to varied critiques. Shelters in particular have been characterised as places of discipline, medicalisation, and control (Gowan, 2010; Hoffman and Coffey, 2008; Lyon-Callo, 2000; Murphy, 2009; Stark, 1994). Wright (1997: 247) argues that shelter providers have also historically engaged in a hostile relationship with street people, perceiving them ‘as not wanting to follow the rules against alcohol or be in at certain hours’. Others, likewise, have stressed that, however controlling such spaces might be, homeless people often resist the given prescriptions or narratives of their abnormality while using services (DeVerteuil, 2014; Lyon-Callo, 2000). Some of this hostility undoubtedly derives from the broader function of welfare policy, namely its mandate to reform recipients into ‘self-sufficient’ individuals, to push people off welfare and out of shelters. How this process unfolds depends upon whether recipients are deemed deserving of care. For those considered unable to work (i.e. ‘disabled’), relief is cast as an entitlement that is deserved. For those ‘able-bodied’, conversely, relief is contingent upon admitting individual fault (bad mindset, addiction) and consenting to a regimen of reformation through which to achieve employment and self-sufficiency. ‘Relief arrangements deal with disorder’, Piven and Cloward (1993: 22) argue, ‘not simply by giving aid to the displaced poor, but by granting it on the condition that they behave in certain ways and, most importantly, on condition that they work’. Though not all spaces of care abide by such conditions, the more established, state-funded organisations, including large homeless shelters, often maintain a focus upon reforming homeless individuals through disciplinary mechanisms. Such reformation vis-à-vis work enforcement, moreover, is codified in broader US welfare policy (Soss et al., 2011).

A purely disinterested care by shelters and other homelessness services, then, is by no means a given. Indeed, the care of such institutions often can be inseparable from the imperative to reform, and to improve the economic position of clientele through individualised prescription and discipline. In addition to the ambivalent nature of care, caring agencies are not immune to the pressures of capitalist growth, and must often compete for funding in an increasingly corporatised non-profit sector (Buckingham, 2009; Cloke et al., 2007a). Further, as DeVerteuil (2014: 880) argues, spaces of care perform the key function of abeyance –‘where the voluntary sector acts as a container’ – and operate within a broader urban landscape that includes punitive measures. As DeVerteuil (2014) suggests, the state ‘cannot implement naked punitiveness without offering up some sort of alternative salvation for those deemed more deserving’. Through our case studies, we show not only how spaces of care distinguish between deserving and undeserving poor people, but how they reinforce the punitive imperatives of revitalisation by actively punishing those deemed undeserving: in this case, the unsheltered homeless.

Several scholars have explored how punitive and compassionate strategies are, in practice, difficult to parse. Johnsen and Fitzpatrick (2010) argue that policing is not always driven by punitive ambitions, and that anti-homeless laws can, ultimately, be put toward compassionate ends. They term this dynamic ‘coercive care’, wherein police ‘safeguard the welfare of those involved in street culture’ through mechanisms of enforcement (2010: 1705). In addition to the worrying politics of this ‘tough love’ paternalism, where the state enforces a notion of wellbeing that trumps any imperative to preserve choice and freedom, Johnsen and Fitzpatrick’s (2010) explanation risks overlooking the broader forces at play. The authors’ focus on the hostility or sympathy of individual actors risks ignoring the economic pressures to remove homeless people to foster greater revenue in a neighbourhood. Further, and despite the best intentions of the police and their charitable counterparts, homeless people, whether seen as ‘self-destructive’ or not, are targeted by anti-homeless laws, fines and potential jail time. Stuart’s (2014, 2016) ethnographic study of policing in Los Angeles’s Skid Row similarly addresses the blending of care and punishment in homelessness management. Focusing on the ‘marginal’ space of Skid Row – as opposed to the ‘prime’ space of central business districts – Stuart (2014: 1911) finds that police officers ‘act primarily as “recovery managers” acting on behalf of rehabilitative organizations to transform homeless people into sober, self-governing and responsible citizens’. In these marginal spaces, the police use existing anti-homeless laws to threaten homeless individuals into the caring spaces of disciplinary reformation.

In the two case studies that follow, we expand on the recognition that criminalisation and compassion operate in tandem. The intent is not to question the presence of genuinely caring relations within homeless geographies (Fresno and Phoenix included), but rather to remain attentive to how such care can operate with punitive, or at least less caring, forms of poverty governance. This goal requires looking beneath the expressed, compassionately framed motives and rationalities of individuals and institutions so as to grasp their material effect. In the Phoenix case study, we illustrate that such ‘coercive care’ not only emerges in marginal spaces, but often encompasses central business districts and the outer suburbs. Fresno sheds light on a new sort of coercive care altogether, with homeless shelters actively lobbying for the destruction of homeless encampments. Both cases, meanwhile, highlight how this blending of paternalistic care and criminalisation derive from a common base: the political-economic force of revitalising cities.

Fresno: Shelters as champions of revanchism

The city of Fresno has long struggled against the widescale presence of poverty and homelessness in its downtown neighbourhoods. In 2011, Fresno was the second most impoverished city in the nation and also experienced the second highest rate of homelessness (Bishaw, 2012; National Alliance to End Homelessness (NAEH), 2012). The downtown area, bounded by the triangular intersection of the city’s three major highways, is one of the poorest in the city. For years, homeless Fresnans took refuge in sprawling encampments concentrated in the historic ‘Chinatown’ district in the southern corner of the downtown, alongside several homeless service facilities. In interviews and casual conversations, Fresnans often lamented that the city has no urban centre for middle-class shopping, recreation, and socialising. As one council member said: ‘In downtown, unfortunately, many of the neighborhoods here are shot. … Business is dried up because they can’t get anybody to come down here anymore’. As such, Fresno officials worked for more than a decade to implement a massive downtown redevelopment project (Speer, 2016, 2017). Yet the downtown encampments were seen to hamper nearby developments, including a US$61 million baseball stadium designed to bolster downtown growth. As such, homeless management in Fresno is intrinsically tied to the city’s perceived need to revitalise urban space.

As part of its revitalisation project, the city passed several laws to move homeless people out of the downtown. Fresno municipal code severely restricts panhandling and imposes prison time on violators (Fresno Municipal Code, 2014 § 9-2608). Pushing a shopping cart is also punishable by US$1000 in fines and up to one year in jail (Fresno Municipal Code, 2014 § 9-31). For more than a decade, the city has also implemented a maverick policy of bulldozing homeless encampments. Over the course of a two-year period beginning in 2005, it conducted at least 50 sweeps. Each time, homeless people built new encampments. By 2006, the city was bulldozing the same encampment in downtown Fresno every two weeks (Kincaid et al. v. Fresno, 2006). In the summer of 2013, it bulldozed every camp in the city and set up a police taskforce to prevent anyone from erecting a tent in the future. Alongside this punitive intervention, the housing authority expanded the provision of housing vouchers to the ‘chronically homeless’ and was widely praised for its homelessness management efforts.

Anti-homeless laws and housing vouchers were both intimately tied to urban revitalisation. In an interview, a homeless man indicated that the heavy presence of homelessness in the downtown area would make revitalisation impossible: ‘They can’t revitalize downtown Fresno … I mean who wants to live like that? Party with a bunch of homeless people?’ When asked what the city should do, he said there was no solution – there were no jobs and not enough shelters, and the city was just going to ‘lock them all up’. Meanwhile, Fresno’s mayor lauded the success of downtown revitalisation, citing the new lofts popping up all over downtown (City of Fresno, 2011). A major downtown property developer explained the connection between homelessness and revitalisation in Fresno:

We need to grow our economy here, so I think the net effect of homelessness here is far more severe than it is in a city that’s already got other assets or attributes which draw people. In my mind, to make revitalization a success, we need to make downtown Fresno a destination. We have not been able, to this point, to identify a workable model for making that connection – make it a destination. And without dealing with the homeless problem I think it simply makes it even more difficult.

This same developer also spearheaded Fresno’s ‘Ten Year Plan to End Chronic Homelessness’ and was instrumental in expanding housing for the homeless while also overseeing encampment evictions. Beyond private businesses, the non-profit sector was also instrumental in shaping homelessness policy. Ensuring their institutional longevity meant that Fresno’s shelters became key players in promoting downtown evictions.

Poverello House is Fresno’s primary homeless service provider, providing free meals three times a day and hosting an outdoor shelter (National Coalition for the Homeless (NCH), 2010). Jim Connell, a former stockbroker, began as CEO in the 1990s to streamline the organisation as it became more institutionalised. Under Connell’s supervision, Poverello encouraged a contentious relationship between shelter staff and street populations and has repeatedly targeted tent cities that develop in front of its property. In 2003, at the request of shelter officials, police officers and sanitation workers dismantled an encampment that had developed on a former junkyard adjacent to the shelter (City of Fresno, 2007). In 2011, police again removed an encampment after the shelter complained, and the shelter banned all campers from using their services (Rhodes, 2011). During court testimony in a case filed against the city seeking restitution for evictions, Connell described the destruction of homeless people’s bicycles and shopping carts as ‘compassionate’ (Rhodes, 2006), suggesting that he viewed his actions as a form of ‘tough love’.

By 2013, the Poverello House was surrounded by three of Fresno’s seven street encampments. Connell explained in a personal interview why he wanted the encampments removed:

We’ve seen the number of people coming for services, particularly at the women’s center and the clinic, go down dramatically this year. And the anecdotal evidence is that particularly the women and children are afraid to go through that gauntlet to get in here. So it is affecting our services. … I’ve had several meetings with the mayor and different people in the administration and they say they plan on cleaning it up.

Connell sought to remove the encampments to ensure that others could more easily access services, and to safeguard his organisation’s attendance rates for funding purposes. In this way, the impetus to care – for both the ‘deserving’ poor and the organisation itself – led to a punitive stance towards homeless campers.

Sherry Oliver ran a day shelter for women located near the Poverello House. Like Connell, she also encouraged evictions and lamented that many of her clientele, mostly the housed poor, had stopped coming since the encampments developed around the property:

You don’t see nearly the families anymore that you used to because I think they just don’t want to come through that maze. … Obviously you don’t want this encampment to be located here. That’s hurting your services and your ability to provide them. … It’s a dilemma, because yes people need their freedom and they have a right to live, but the impact on everybody else is very difficult.

She was worried that the shelter’s summer programmes might be cancelled because of low attendance, and sympathised with those who did not want to walk through the encampments. Thus, she saw that the very existence of her organisation was threatened by the presence of encampments. Just as private businesses often complained that homelessness prevented customers from patronising their stores, Fresno shelters also sought to protect their attendance rates by removing encampments outside their facilities. In the process, they were forced into the paradoxical position of simultaneously caring for homeless people while championing their forced removal from public space.

In addition to protecting their attendance rates, shelter operators were also concerned with government approval. With Fresno’s downtown being encroached by development, shelters felt the squeeze of revitalisation. Faced with the ongoing threat of city opposition, shelter operators repeatedly complained that their existence was tenuous. City officials often framed downtown shelters as a magnet for the homeless and occasionally threatened to remove the shelters altogether. This magnetism exacerbated the hostile relationship between shelter operators and street campers. As a local activist recounted, he had seen ‘constant proposals for moving the Poverello and the Rescue Mission even further south of the downtown area’. When Connell was asked if he was worried about the city’s push to revitalise downtown, he said, ‘That’s the city’s attitude – the reason we have the problem out there is because we’re here. Bruce Rudd, the city manager, that’s his opinion. They [the homeless] wouldn’t be there if we weren’t here’. For Connell, the concern was not simply that his numbers were declining but also that the camps threatened the continued existence of his organisation.

The Fresno Rescue Mission, a local religious shelter for men, was headed for years by Larry Arce, a former reverend and probation officer. Over the years, the Mission had been periodically threatened with city disapproval, and Arce had taken matters into his own hands by directing and funding a series of private evictions. In 2006, Mission staff, armed with a fork lift and dumpster, destroyed a nearby encampment on public property. One woman lost her tent and had nowhere to sleep that night. When activists tried to retrieve the tent from the Mission’s dumpster, Arce called the police and accused them of stealing (Rhodes, 2007). In 2008, several homeless Fresnans sued the Mission, describing in their complaint how the Mission destroyed tents containing bedding, clothing, jewellery, family photos, birth certificates, and even heart medicine. During court testimony, Arce admitted that the shelter destroyed people’s property even when they left their carts on the street to stand in line at the Mission for a meal (Rhodes, 2008). Yet Arce’s actions cannot be explained only as personal animus. In an interview, he described experiencing pressure from the city:

Years ago they wanted to move us from here because they were building the Chukchansi stadium. … But they were saying that if we’re going to have this we need to get rid of all the homeless. We can’t have them here. They’ll be out here in the front. They’ll be accosting people. They’ll be begging. All kinds of perceived things were going to happen so they were blaming us – that we had to leave, we couldn’t be here, we had to get out of the way.

Thus, Arce’s aggressive stance against encampments was rooted in the bottom line of his organisation – if the city saw that his shelter was surrounded by street encampments, it would be targeted for removal as revitalisation expanded into the neighbourhood.

Aside from the pressures of revitalisation, notions of care themselves were often rooted in capitalist ideologies that distinguish between the ‘deserving’ and ‘undeserving’ poor. Like many in Fresno, Oliver perceived the homeless people living in encampments as hardened criminals. She said: ‘You take your life in your hands to walk down F Street right now. … I’m not sure where some of these people came from’. She told me she had been lobbying the city to remove the encampments for quite some time and said, ‘hopefully, things will move now that they’ve gotten the message that it is violent out there’. Arce’s actions were also deeply informed by an evangelical vision of care as religious reformation. In 2007, he had been the driving force behind a proposal to officially criminalise camping city-wide. He said:

Me and Jim Connell worked on a comprehensive project … where the city would pass an ordinance and say you cannot camp out willy nilly anywhere you want to, in front of people’s houses, in front of businesses out on the streets. … That’s not the way people live. You need to get into a program, you need to get your life together.

Thus, Arce promoted evictions and criminalisation because he rejected ‘the way people live’ in the encampments. He was relieved when Fresno’s Police Chief, who was also a Rescue Mission board member, began pushing aggressively for evictions in 2013.

As such, encouragement of evictions is not evidence of a lack of care, but of the ambivalent nature of care and its vulnerability to the pressures of revitalisation. Although shelter operators urged the city to bulldoze the encampments surrounding their shelters, they simultaneously described themselves as having compassion for the poor. Yet the nature of their organisations – as private institutions in need of continued funding and government support – demanded punitive action. As Cloke et al. (2007a) note, the individual agency of volunteers is often impinged by organisational practices and norms. When private institutions are tasked with caring for the homeless, their need to attend to their own financial security often requires punitive action. The situation in Fresno reveals how private institutions of care are also subject to the pressures and ideologies of capitalism that punish homeless people, and in their efforts to ‘care’ for the institution itself, as well as its residents, managers furthered punitive practices against homeless people who live in public.

Phoenix: Abeyance as punishment

In 2005, Phoenix developed ‘The Human Services Campus’, an enclosed space designed to consolidate and streamline homelessness-related services. It became a cause for celebration. ‘Until now’, William Hermann (2005), columnist for the Arizona Republic, wrote, ‘the chronically homeless, many suffering from mental illness, have had to go to many different agencies to find help, [but] today they are all on one campus’. ‘The $24 million, 14-acre campus’, he continued, ‘brings together five agencies that help the homeless, as well as therapists, counselors, and advisers from state, county, city, and private agencies’. The coerced emplacement of the Campus around CASS (Central Arizona Shelter Services), the metro region’s largest shelter, solidified the impoverished neighbourhoods of south Phoenix as the place to access supportive services (Brinegar, 2003). Present for the grand opening, a mental health and homelessness advisor to the Bush Administration opined that the Campus ‘could be a national model’ (Hermann, 2005). Since 2005, the facility continually attracts 1300 individuals each day (Arizona Republic, 2010) and, in 2012, sheltered 740 nightly (HUD, 2012) of the estimated 25,832 people experiencing homelessness throughout the year (Department of Economic Security (DES), 2012). The Campus, Hermann concluded, ‘is more than a haven, it’s a magnet’.

Indeed, the executive director of CASS, Mark Hollernan, stressed the project’s ingenuity in terms of its geographical pull. ‘We’ve learned you need to offer a sort of “one-stop” shopping approach,’ he explained. ‘You provide a comfortable place to get off the streets, provide food, medical care, substance abuse treatment and mental-health counseling, housing advisers. … Everything and everybody in one place’ (quoted in Hermann, 2005). Beyond pure, disinterested care, however, later reports explicitly state what Hollernan only implies: that the Campus also functions to protect adjacent property values and keep homeless individuals from walking a mile northward to gentrifying downtown Phoenix. ‘Our first outcome’, Hollernan explains, ‘is to get everybody [homeless people] there [to the shelter] tomorrow, nothing else, because if we know they’re there, that means they’re not somewhere else potentially causing problems to themselves, the people around them or properties’ (quoted in Lee, 2013). Here, it is clear that the Campus not only embodies a collection of relief services dedicated to caring for homeless people, as a ‘space of care’, it also explicitly functions to remove homeless people from public space, away from the nearby, gentrifying downtown. In this way, its goal mirrors that of anti-homeless laws and anti-homeless architecture despite sharp differences in the given rationales (e.g. criminalisation versus compassion).

The ‘magnetism’ of the Campus, though, is not complete. Its incompleteness has much to do with the Campus’s surrounding geography. The south Phoenix neighbourhood that houses the Campus is, in street lexicon, referred to as ‘The Twilight Zone’ (‘The Zone’ for short) in reference to its high crime rate, rampant drug use, homelessness and prostitution, factors that preceded and are now exacerbated by the Campus’s presence (Amster, 2008; Brinegar, 2003). Indeed, the Zone is ‘the wrong side of the tracks’ epitomised, a place of danger and heterotopic decay only to be ventured toward in cases of desperation (when no other areas provided needed services). ‘A single man trying to get into shelter’, Theresa James, former Homelessness Coordinator for the nearby City of Tempe, explained, ‘has two choices: CASS and the men’s shelter in Mesa [another metro suburb], which only permits the employed. Frankly’, she continued, ‘I wouldn’t want to go down to CASS either’. Sam, a formerly homeless man, likewise expressed the dangers of the Zone: ‘I don’t go down by the [railroad] tracks by CASS or stuff because you’ll get killed. That’s just the way it is’. These observations reveal how the Campus is, for many, a place of last resort. It also attests to why the south-eastern suburb of Tempe, a city with no permanent shelter and only a handful of services, contains more than its proportional share of homeless individuals – once estimated at around 500 (Hermann, 2008) – even with a substantial arsenal of anti-homeless ordinances.

The magnetism of the Campus, then, is not strong enough to contain homeless individuals, despite its vast, superstore-esque expansion and its potential to be a national model of homeless relief services. It is within this incapacity that other modes of spatially managing homeless people come into play, particularly anti-homeless laws. Indeed, criminalisation and compassion have always worked together throughout greater Phoenix’s history of homeless management. Several years after grabbing national headlines for outlawing lying down or sleeping in public space, banning relief services in its erstwhile skid row, and then bulldozing a large homeless encampment over the 1982 Thanksgiving holiday (Alter et al., 1984; Brinegar, 2003; Luckingham, 1989; Sexton, 1983), Phoenix established CASS, the city’s first large-scale permanent shelter and predecessor to the Campus. In 2005, the same year CASS was upgraded to the Campus, the Phoenix City Council passed its current anti-camping ordinance, joining the other citywide ordinances that criminalise ‘aggressive panhandling’ and loitering. The timing of these interventions – simultaneously increasing shelter capacity and anti-homeless policing – suggests that they work jointly to spatially manage homeless populations.

Crackdowns on homeless people and their means of survival, then, remained both significantly ineffective and, in the eyes of business improvement districts and policymakers, perpetually necessary. Even after metro Phoenix became a leader in the compassionate management of homelessness with the Campus, it continually swept homeless people from parks (Poletta, 2011), evicted and bulldozed riverbed encampments (Coe, 2011), installed anti-homeless architecture (Walsh, 2012), and stepped up anti-panhandling campaigns (Smokey, 2008). Tempe, for instance, undertook a four-day ‘sting operation’ on Mill Avenue central business district in 2009, complete with undercover police officers, to crack down on aggressive panhandling (Náñez, 2009a). Although each of these cases deployed older laws, in each instance journalists attached a compassionate framing to these punitive tactics. Upon arrest or eviction, newspaper articles explain how homeless individuals were compassionately counselled by regretful police officers about the resources offered by the Human Services Campus. During Tempe’s crackdown, local police distributed cards containing both information about services and the specifics of the anti-panhandling law (Náñez, 2009b), such that criminalisation and compassion were all on a single pocket-able card. In this sense, Tempe police acted in a ‘caring’ capacity, offering on-the-street advice on where to receive care, at the very same time they presented the threat of imminent arrest.

Another example of the blurring between the pull of services and the push of anti-homeless laws is captured by recent police actions in the Sunnyslope neighbourhood of north-central Phoenix, a poorer suburb with a disproportionate number of homeless individuals. Phoenix police have continually targeted ‘homeless hangouts’ through monthly sweeps as residents and businesses repeatedly report ‘transients drifting into commercial areas after stopping at nearby homeless shelters and soup kitchens’ (emphasis added) (Ferraresi, 2009b). Homeless populations in Sunnyslope are, in this example, posed as objects aimlessly moving about the community and ‘drifting’ into foreign spaces. But their infraction is framed as both deviant and pitiful, and the actions of the police reflect this framing. ‘The first thing officers do on sweeps’, Ferraresi (2009a) explains, ‘is to provide transients with information on homeless shelters and human services,’ while the second thing they do, the article states, ‘is push for maximum 180-day jail sentences for transients in violation of travel restrictions [trespassing stipulations]’. One could argue that these officers first provided service referrals to skirt claims of cruelty, but it is also likely that the Phoenix Police Department comprehends the limitations of purely punitive campaigns to spatially manage homeless populations.

The intertwined processes of care and punishment are even more apparent around the marginal spaces of the Zone, as Stuart (2014) found in Los Angeles. ‘They act like cops but think like social workers,’ a column in The Arizona Republic begins. ‘About seven years ago [2005]’, columnist Michelle Lee (2012) explains, ‘the Phoenix Police Department walking-beat squad that patrols the Human Services Campus in downtown Phoenix came up with a new strategy to deal with homeless people on their beat: help them connect with social services rather than arresting them for every minor misdemeanor’. ‘There was a philosophical change,’ Lieutenant Sean Connolly of the Phoenix Police Department recalled. ‘All we would do is arrest our way to solutions. There was no sustainability in the process’ (quoted in Lee, 2012). The ‘solutions’ of Phoenix’s anti-homeless laws, in other words, significantly failed to erase homelessness while likely costing the city fortunes in booking and jailing (see NLCHP, 2011). Disillusioned, in 2005 officers patrolling the Zone were given special training in crisis intervention and social work interviewing techniques to better manage the ‘chronically homeless’ and mentally ill. Under this new strategy, immediate arrest is no longer the sole action used to clear homeless people from the street and into the Human Services Campus. Instead, officers ‘use the threat of arrest as leverage to keep them on track with their services, making arrests only if they have exhausted all other options’, a strategy that reduces punitive costs and helps push the otherwise-intractable homeless population into the Campus or related agencies for shelter and rehabilitative services. But the mere threat of arrest does not always work to move homeless people along or toward services. Although these police officers have special training such that they can mimic the role of a social worker, Lieutenant Connolly stresses that arrests can be, and often are, made on their homeless beat. ‘They make it clear,’ Lee (2012) concludes, ‘“Don’t take my kindness for weakness”’.

Even when these officers’ ‘kindness’ is not mistaken for weakness, the Human Services Campus, along with police officers and businesses in the adjacent central business district, still find it difficult to contain homeless people in the Campus and the depressed neighbourhood that hosts this institutional enclave. In 2006, in an effort to rectify this apparent shortcoming, CASS developed a programme whereby ‘a team of Phoenix police and CASS specialists began crisscrossing downtown Phoenix four nights a week to engage homeless people who have severed connections with service providers’ (Collom, 2009). Night ‘outreach’ is accompanied by daybreak patrols that wake homeless individuals they find and then attempt to ‘persuade them to sign up for the services they need’ (Lee, 2012). This melding of social work and penal strategies aims at moving homeless populations off the streets and into shelters (through threatening arrest, offering referrals in addition to arrest, or doing street outreach and transporting individuals to the Campus). Even with such punitive and compassionate tenacity, the outreach team reports that only 35% of those confronted are persuaded and enter into some sort of ‘treatment’ (Collom, 2009), which could entail anything from entering drug rehab to meeting with a case manager. The pull of treatment/shelter and the push of anti-homeless laws remains considerably ineffective, such that push and pull factors in the geographies of homelessness often function together to drag homeless people into services, and away from public spaces.

Rather than primarily enacting ‘compassionate’ policies throughout the 1980s ‘homeless crisis’, as some general histories of American homelessness suggest (Gowan, 2010; Mitchell, 2011), from the beginning metropolitan Phoenix combined punitive and compassionate forms of management to constrict the geography of homeless people. Indeed, these two modes of managing the homeless continually compensate for the other’s failings, pushing where the shelter’s pull was insufficient, pulling (and concentrating) when anti-homeless laws had aimlessly banished. Such blending of criminalisation and compassion has evolved to the point where they are in fact indistinguishable, where ‘cops think like social workers’ and social workers walk the beat. This, in turn, shows how the city’s geography of care – despite being lauded as a national model – is crucial to the project of excluding homeless people from the city in order to create a more palatable urban landscape.

Conclusion

These two case studies reveal that binary analyses of care and punishment distract from more dialectical understandings of how revitalisation impacts those excluded from the protections of private property. In practice, the city is neither a flat field of punitive processes, nor a terrain marked by the proliferation of private caring spaces. Further, although both care and punishment have different means and represent separate ideologies, the ends can sometimes be the same: to spatially manage poor and homeless populations to shore-up capital projects in revitalising urban spaces. In both Fresno and Phoenix, homeless shelters assisted in the project of criminal enforcement. Simultaneously, anti-homeless laws were framed as compassionate mechanisms by which to push homeless people into the spaces of care they so desperately needed. Thus, compassion and criminalisation must be understood recursively, as the line between social worker and police officer is intentionally muddled in Phoenix, and shelter operators in Fresno become architects of eviction.

Beyond the blurred geography of care and punishment, this paper shows how institutions of care are themselves frequently subject to the revanchist pressures of urbanisation. As our case studies reveal, such institutions are part of a larger programme of homelessness management that must negotiate the capitalist city. In Fresno and Phoenix, private caring institutions and actors often responded to revitalisation by promoting punitive practices against the unsheltered homeless. In turn, unsheltered homeless people were often deemed ‘undeserving’ of care for refusing to be sheltered. This is not to say that all caring organisations are problematic or intrinsically complicit in punitive actions. Indeed, both Fresno and Phoenix are home to activist movements that promote the rights of homeless people, as well as care workers motivated by a deep commitment to empathy, equality and shared humanity. As others have found, genuinely caring spaces do exist. Yet rather than upholding care as universally pure or beneficial, we urge that urban studies must examine the pressures and constraints that can produce a compromised geography of care.

This is not only an abstract or theoretical intervention into the literature on urban homelessness. In analysing shelters as politically and economically situated rather than independent spaces of care, scholars can remain sensitive to the ways in which private institutions of homelessness relief can be deeply imbricated with the punitive state apparatus, and how homelessness is addressed by a wide range of institutions – shelters, jails, welfare offices, and prisons, for example – that all impose varying degrees of socio-spatial control, caring or otherwise. In the contemporary US city, the capitalist development of urban space still functions to control homeless people’s movements, regardless of whether this control is motivated by compassion or revanchism. Thus, it is essential to examine on-the-ground spatial dynamics of homelessness management holistically rather than characterising spaces of care as discrete sites that operate outside of the pressures of the capitalist city. As cities increasingly unroll new caring models of homelessness management and open new spaces of care, these policies and spaces must remain subject to critical examination and placed in the larger context of urban capitalism. With this in mind, both traditionally ‘caring’ and ‘punitive’ institutions must seek to overcome the institutional and ideological barriers to achieving a more radical, inclusive, and empowering city for everyone. The project must not only include the pitiful right to stay put on the street (away from services and out of jail), but also the right to quality housing, income, and an overall dignified life.