Boom,bust,churn: Prison closure and prison expansion in New South Wales,Australia

Abstract

This paper argues that rather than signalling contraction, prison closures can operate to restructure prison systems for further growth. Focusing on New South Wales (NSW), Australia, we analyse the closure and downsizing of four prisons in 2011–2012 and their relationship to a $3.8 billion prison construction programme launched just a few years later. We draw on Gilmore’s concept of the carceral ‘spatial fix’ to show how closures reorganised the state’s prison system by transforming surplus carceral capacity into fiscal capacity and creating political and economic space for new mega-prisons. The paper develops the concept of ‘prison churn’—the rapid and cyclical reworking of prison infrastructure—as a temporal mechanism through which the prison fix is operationalised. We trace the prison churn in NSW, which included prison closure, downsizing, reopening, redevelopment, and decommissioning through case studies of four prisons. The case studies illustrate how each decision taken by the state was shaped by competing, local priorities such as urban rent extraction and the desires for cheap prison labour and employment for prison staff in the regions. Our analysis reframes infrastructural restructuring as a crisis management tool that entrenches, rather than diminishes, carceral state power.

Introduction

Across the settler states of the United States, Canada, and Australia, the late 20th and early 21st centuries saw an unprecedented expansion of carceral infrastructure that includes a wide range of prison types (from jails, through minimum security facilities, and all the way to supermax prisons), alongside non-prison infrastructure that furthers carceral state power (see, for instance: Morrell, 2025; Norton et al., 2024; Pelot-Hobbs, 2023; Piché et al., 2017; Russell, 2024; Story, 2019). In the United States, the number of state and federal prisons more than doubled between the 1970s and 2000s, and county jails have since mushroomed as a new site of incarceration. In Australia, which remains under researched in carceral and abolition geographies (but see Russell, 2024), the national imprisonment rate rose steadily from the 1980s onward (Leigh, 2020), with particular acceleration in the post-2000 period as jurisdictions such as Victoria and New South Wales undertook large-scale prison construction programmes (Russell et al., 2025).

This paper examines the context for and events that preceded the prison boom in New South Wales, which saw an expansion of the system by over 6000 beds between 2016 and 2022 alone. We show that this boom was constituted, in part, through a ‘prison bust’—the closure and downsizing of four prisons across the state between 2011 and 2012. Rather than a linear trajectory of expansion, the ‘bust’ reveals the relational and spatially uneven dynamics of carceral development. We argue that prison closures were not indicative of decarceration but of a reorganisation of state carceral power—constitutive of a prison fix through which the state consolidated and expanded its punitive infrastructure under the guise of economic rationality. These closures displaced carceral functions and populations, facilitating the construction of new, larger prisons in regional areas under neoliberal development logics. While restructuring must be understood within the ongoing project of settler colonial governance, where prison building has long been a technique for managing racialised surplus populations, it is also a tool of local economic intervention. In this way, the NSW prison ‘bust’ of 2011–2012 was not an undoing of carceral power but an integral moment in its reproduction.

The paper illustrates that the process of prison expansion is not unidirectional and can include dynamics that appear to be shrinking the system, but in fact do not. In this sense, prison closures can be ‘constitutive of carceral expansion’ (Giannacopoulos and Loughnan, 2020). We show that the broader infrastructural agenda of prison expansion in NSW included the closure of prisons at different stages, and that this was not an aberration. Rather, prison closures are an internal dynamic of the process that facilitated NSW’s transition from a system relying on 19th century carceral facilities to a ‘modernised’ one. In that sense, the wave of prison expansion we document in NSW is rooted in the long history of NSW as a penal colony. Thus, to understand how the state grew its prison infrastructure since 2016, we must consider the conjuncture within which this process was shaped—not as response to ‘new’ crisis but an extension of the state’s foundational logic of carceral colonialism (McKinnon, 2020). This logic continues to inform how NSW functions today, applying punitive, extractive, and dispossessory mechanisms to maintain a racial capitalist order while incarcerating Indigenous people at staggering rates. ¹

Through a series of case studies, we describe below the economic motives animating carceral planning decisions in NSW. In one moment, the closure and downsizing of old prisons increase the state’s financial capacity, transforming bedspace ‘surplus capacity’ to fiscal capacity by reducing the cost of maintaining prisoners in so-called ‘inefficient’ and ‘outdated’ prisons. Then, in deciding to reopen some of these older prisons, if only temporarily, we see that the state was able to create not just ‘bedspace’ but also much sought after demand for labour (of both staff and inmates). The diverse responses to prison reopening illustrate that for rural/regional ‘host communities’, not just corrective services labour but also prisoners’ labour was desired. In a third moment, a different economic dynamic played out—the reopening of a prison in metropolitan Sydney was bitterly opposed and eventually prevented since the value to be gained by permanently closing the prison far exceeded what was to be gained by its reopening. In this case, the prison was located in an area of substantial urban development and financial investment that made its reopening risky and undesirable. Here, neither staff employment nor cheap prisoners’ labour could compete with the value to be extracted from land appreciation, which necessitated that the prison remained closed. Taken together, the case studies illustrate how prison expansion took place through a cyclical and fast-moving restructuring that included closure, downsizing, mothballing, reopening, decommissioning, and, eventually, selling off of significant prison assets. Building on Russell et al. (2022), we refer to this restructuring as a ‘prison churn’, a term which helps us draw out, first, the contigent and context-specific nature of carceral restructuring. Namely, the way that the churn manifests differently across diverse geographies given their local political and economic processes and dynamics. Second, the ‘prison churn’ helps expose the mechanism throughout which the carceral spatial fix is operationalised (following, Gimore, 2007), and its specific temporalities.

The paper expands existing literature on the political economy of prison geographies in several ways. First, it documents what prison restructuring looks like from Australia, which remains vastly under researched in the field. Beyond empirical novelty, the paper sheds new theoretical light on what recent US-based scholars identify as a ‘prison bust’ (Harris et al., 2024) or ‘prison devolution’ (Norton et al., 2024), denoting the vast closure of state prisons and concentration of incarceration in county jails. While recognised as a form of restructuring of carceral state power, rather than a mode of decarceration, prison closures in the US have not been analysed in similar terms to the prison boom—in Gilmore’s (2007) words, as a fix. Examining prison closures from NSW, we show, first, that prison closures (temporary or permanent) are endemic to state carceral expansion; second, that prison closures are consistent with the logic of the spatial fix; and finally, that the reorganisation of prison infrastructure is tethered to the longue durée of settler colonial carcerality, a context that is often marginalised in the US-centric literature, despite its obvious relevance.

Methodologucally, the paper offers a description of medium-term prison statistics in NSW, before undertaking a qualitiative analsysis of four case studies, Berrima, Parramatta, Kirkconnell, and Grafton. These prisons were selected since they were all closed or substantially downsized within a narrow temporal window (2011–2012) but subsequently followed divergent trajectories of reopening, redevelopment, or permanent closure. This bounded selection allows us to hold the broader policy and fiscal context constant while tracing how distinct local political economies of each case study such as urban land markets, regional labour dependence, and infrastructural obsolescence, shaped prison planning. Conjunctural analysis is operationalised here by situating each case at the intersection of short-term policy interventions (such as bail reform, budgetary rationalisation, and emergency responses to overcrowding) and longer-term structural dynamics, including settler colonial carcerality, neoliberal infrastructure reform, and spatial strategies of state crisis management. The study is limited by its reliance on publicly available documents and does not include ethnographic or interview-based data with incarcerated people, correctional staff, or policymakers. The paper does not aim to provide a comprehensive inventory of prison restructuring across NSW. Instead, the four cases are used analytically to examinate the mechanisms of the prison churn and to theorise how infrastructural instability functions as a constitutive element of the prison fix.

The paper proceeds as follows. The literature review revisits Gilmore’s concept of the carceral spatial fix, arguing that prison closures are integral to, rather than exceptions from, prison expansion. We develop the term ‘prison churn’ and clarify that while ‘prison fix’ alludes to a long-term restructuring of prison geographies, prison churn signifies the swift and cyclical means by which the carceral state expands in the face of and response to converging crises. The section, ‘carcerality, coloniality, and ‘Better Prisons’ in NSW’ provides context about the NSW prison system, placing events from 2010 onwards in the longer history of NSW’s colonial carcerality. [In] the empirical section we present case studies of four prisons that illustrate the localised workings of the prison churn while bringing the different cases into conversation with each other to explicate a broader trend. We conclude by highlighting the contribution of non-US prison research to the study of carceral restructuring, and the need for further regional, conjunctural analyses of Australia’s unprecedented prison expansion.

Literature review

Re-reading the prison fix

Following Gilmore (1999, 2007), prison expansion has been increasingly understood through the lens of the spatial fix (e.g. Brooks and Best, 2021; Morrell, 2012; Russell, 2024). Harvey (2018) describes the spatial fix as a mechanism by which capitalism attempts to manage its inherent and internally driven crisis tendencies through geographical expansion and spatial reorganisation. This manifests in two primary ways (Jessop, 2006): attempts to manage crises by anchoring surplus capital in fixed infrastructures, and by extending into new markets. Harvey argues that both strategies displace rather than resolve capitalism’s contradictions. Thus, the spatial fix is inherently unstable and temporary, often leading to the intensification of capitalist crises, and therefore necessitating a perpetual cycle of destructive restructuring across the landscape.

Following Harvey, Gilmore (2007) argues that prison expansion in the United States forms a carceral spatial fix, or a ‘prison fix’. Prison expansion is not, she argues, a neutral infrastructural decision, but a state strategy to provisionally manage the crises of racial capitalism. Gilmore shows that the expansion of carceral infrastructure serves as a geographical, economic, and political response to surpluses of land, labour, capital, and state capacity. These surpluses are themselves produced through processes of settler colonialism, deindustrialisation, environmental degradation, and state retrenchment (Russell, 2024; Schept, 2022). If mass incarceration, according to Gilmore, is a form of class warfare, then the prison fix is a state tactic for managing that war. Yet, as Harvey makes clear in his framing, the prison fix is a mechanism by which the state displaces rather than resolves its crises. The prison fix highlights that dominant explanations for increased incarceration—like moral panic over crime, drug epidemics, or structural changes in employment—while relevant, fail to answer the key questions, in Gilmore’s words, of ‘who is being punished, for what, and to what end?’ (Gilmore, 1999: 173). Such explanations ultimately provide ‘the ideological legitimacy’ (Gilmore, 2007: 114) for economic restructuring. The prison fix thus serves to manage social and economic crises while magnifying state power and capacity.

Taking cues from Gilmore, Morrell (2025) argues that the expansion of prisons in Upstate New York State between 1971 and 2000, including the construction of 39 new facilities, was conceived of and implemented as a political and economic project designed to counteract (predominantly white) working class job loss, resulting from post-Fordist economic restructuring. Prisons were viewed as a solution to a range of significant social, political, and economic problems, particularly in poor and rural areas, and as a mechanism for capital infusion into distressed parts of the state. Furthermore, prisons helped reinforce racial differences in the workforce—taking existing labour structures from factories into the new prison industry—thus maintaining the property of whiteness, especially in towns that were severely hit by the decline in productive economic activity.

Reflecting on the decrease of New York’s prison population, from nearly 70,000 in 1999 to 32,750 in 2024, Morrell (2025) argues that ‘carceral reindustrialisation’ failed as a strategy of economic development. What appeared in her earlier (Morrell, 2012) work as a reconfiguration of incarceration towards fewer but harsher maximum-security sites is revealed by 2025 as a form of carceral restructuring. Even flagship maximum-security prisons, such as the Southport Correctional Facility, were shuttered. The closure of these institutions exposed as a fiction the proposition that prisons could offer a durable economic solution for deindustrialised rural regions. Yet Morrell (2025) insists that this failure does not signal a retreat of carceral power. The prison, she argues, is never truly a place set apart. Its geographies are porous, leaking institutional violence and social relations into everyday community life. Carceral state-making thus persists through diffuse, racialised practices, from the ‘violence work’ of correctional officers that follows them home, to the routine policing of residents that mobilises racial panic to naturalise Blackness as criminal.

Morrell’s (2012) earlier work foreshadowed recent prison research on shifts in U.S. incarceration. These works show that the expansion of carceral state power is not linear or pre-determined, as the system continually restructures (Norton et al., 2024; Pelot-Hobbs, 2023). In Louisiana, for instance, the vast project of prison expansion was carried out amidst persistent community protest which compelled the state to reorganise its infrastructure of incarceration, displacing prisoners from large state prisons to smaller county jails (Pelot-Hobbs, 2023). Norton et al. (2024: 7) show that this is a US-wide trend of ‘carceral devolution’ in which thousands of state prisons are being shut down or downsized. According to their study, the bedspace capacity of state prisons across the US was reduced by 81,444 between 2000 and 2022 alone; however, an increasing number of people are being held in local jails as a replacement for state facilities. Echoing Morrell, they argue that this shift has not diminished carceral state power but rather diversified and redirected its economic benefits towards sheriff offices who are remunerated generously for taking on carceral executive roles. Jails, as cheaper means of imprisonment, provide a new way of managing surplus land, labour, and populations in the ongoing race war Gilmore describes.

In their careful documentation of the US ‘prison bust’, Harris et al. (2024) show that since around 2000, prison closures have gradually outpaced prison constructions. Adding another dimension to the analysis of this trend, Harris et al. explain that the prison ‘bust’ is a direct consequence of the ‘boom’ during the period 1970–2000. During the boom, states constructed prisons, particularly in rural areas, with the prevailing assumption that these facilities would stimulate struggling local economies through job creation, local contracts, and visitor patronage. However, Harris et al. note that these gains were often illusory, at best limited to public employment and to the short-term, with many jobs filled by residents from neighbouring towns. Thus, prisons did not provide the promised panacea for economic woes but led to an oversupply of prison beds.

Beyond documenting these changes, Harris et al. (2024) provide little guidance on the meaning of the prison bust. Instead, they claim that the future of the prison bust remains uncertain, outlining three possible trajectories: continuation, stagnation, and reversal. The re-emergence of punitive law-and-order discourse signals the potential for reversal, echoing carceral expansions of the late 20th century. Conversely, the authors note that the bust has been driven by fiscal pressures rather than abolitionist politics, raising the possibility of stagnation once states have trimmed perceived ‘carceral excess’. Finally, they argue that while pandemic-era reforms reduced prison populations and infrastructure demand, it is unclear whether these shifts will endure. Harris et al. conclude that while a prison bust may be desired by abolitionists advocating reduced reliance on punitive infrastructure, prison closures do not necessarily represent abolitionist outcomes.

Our analysis of prison closures and reopenings in NSW between 2011 and 2015 builds on and extends the above research. We see that prison closures in NSW in this period represent not the diminishing of carceral state power, but its reorganisation and expansion. Unlike the US case, though, NSW has seen a consolidation rather than devolution of incarceration, with an overall trajectory that similarly represents the widening of punitive capacities. We document and explain this trend later in the paper, but first we clarify why the restructuring witnessed in NSW should be understood as a spatial fix. In so doing, we foreground the argument—intimated but not explicitly made in the above research—that prison closures (and not just prison openings) can also be a form of what Gilmore calls ‘the prison fix’.

Prison closures as a fix

In this paper we claim that prison closures do not necessarily mark an end of the prison boom, or of state institutional contraction. Rather than signifying the rollback of the carceral state, we argue that prison closures can be a constitutive moment within the rollout of the carceral spatial fix—a mechanism through which the prison system manages its internal contradictions by reorganising space and infrastructure. As we show in the following section, the churn of prison infrastructure in NSW—including closure or downsizing, reopening, expansion, redevelopment, and decommissioning—functions as a mode of crisis management, not a sign of resolution. ²

Borrowing the term ‘carceral churn’ coined by Russell et al. (2022) to refer to the moving of criminalised subjects through bail and remand courts, we illustrate in this paper that the churning of prison infrastructure—rather than of people—is a tool for managing economic and political crises while reproducing and expanding the state’s capacity to incarcerate. In this case, ‘prison churn’ represents the shift in modality from an open to a closed prison, and vice versa, from closed facility to a reopened one. This churn mirrors the broader logics of carceral governance, in which populations are continuously moved, reclassified, and warehoused in response to the shifting demands of the state (see also Brooks and Best, 2021). Where the carceral churn reproduces criminality through movement and reprocessing (Russell et al., 2022), the prison churn reproduces spatial capacity—bed spaces—through the ongoing recalibration of carceral infrastructure.

The paper develops the concept of prison churn not as an alternative to the prison fix but as one of its constitutive mechanisms. The prison fix describes the underlying logic of carceral expansion, the long-term displacement of crises of racial capitalism, and settler colonial governance through investment in punitive infrastructure. Prison churn, by contrast, captures the temporal mechanism through which this logic is operationalised. It refers to the cyclical reworking of prison infrastructure, its closure, downsizing, reopening, redevelopment, and decommissioning. It is through this sequence that the state manages the instabilities that carceral expansion produces. In this sense, churn is a form of fix, in that it is a short-term, iterative strategy that stabilises the prison system at moments of crisis and, in doing so, feeds into and sustains the longer-term prison fix. Where the prison fix unfolds over decades and across regions, prison churn operates over shorter time horizons, responding to conjunctural pressures while preparing the ground for further expansion.

Our contribution is to make this dynamic legible by naming the mechanism through which the prison fix is enacted. We argue that the fix does not operate only through expansion understood as new construction, but also through churn. Prison churn is a means through which the fix takes place. Churn stabilises the system at moments of overcrowding, budgetary stress, or political backlash, not by undoing carcerality but by redistributing it across space and time and by converting prison resources into fiscal resources (or the converse). In this sense, churn does the work of ‘fixing’. It is the practical, iterative restructuring through which the prison fix is continually reproduced. The analytical value of prison churn lies in clarifying how apparently contradictory moments of contraction are not deviations from carceral growth, but the means by which the prison fix is continually reproduced.

We read prison restructuring through a conjunctural lens (Hall and Massey, 2010). Thus, we do not ask whether closures represent progressive or regressive policymaking. Instead, our research asks: what work does prison churn do to reproduce the prison system under conditions of instability? Complicating matters further is the fact that prison restructuring is animated by pressures that emerge from within the political economy of incarceration. On one side, we observe drivers that tend toward contraction. Legal reforms such as progressive bail reform can temporarily reduce prison populations, as does the falling incidence of crime. This alleviates some of the capacity and political pressures on the system. Humanitarian or infrastructural reform agendas then frame closures as a moral imperative, especially when older facilities are deemed unfit or inhumane. This reformist impulse is deeply entangled with investment in new carceral infrastructures that increases the ‘efficiency’ of carceral labour-power, replacing labour with capital. Under the guise of efficiency, larger prisons warehouse an increasing number of people with fewer staff, justifying the abandonment of older facilities in favour of technologically advanced mega-prisons with improved ‘natural surveillance’.

Yet carceral infrastructures do not expand smoothly or evenly. The decisions made by the state at one moment—for example to close a prison in response to state fiscal crisis—may generate new crises at another: overcrowding, political backlash, or other forms of infrastructural shortfall. This new crisis requires further interventions, leading to new construction or the reopening of previously shuttered sites. Thus, we claim that what appears from one vantage point as disorder is in fact the carceral state’s attempt to manage unresolvable crises through spatial and institutional reorganisation. In this analysis, the prison churn is not a sign of carceral decline. Rather, it is a mechanism through which the prison system reproduces itself—absorbing critique, adapting to crisis, and reasserting control. In what follows we delineate the conditions, policies, and mechanisms for the restructuring of the NSW prison system over the past two decades, situating them in the history of this colony.

By focusing on the state as the key actor in the restructuring of carceral infrastructure, we do not deny other forces impact the prison churn (e.g. property development and local economies). However, our focus is on the workings of the state and its agents, which is a crucial lens for the study of carceral geographies. As Morrell (2025: 6) poignantly puts it: ‘To study the prison is to study the state: its breath and power, its structures and bureaucracies, its arsenal, its methods of punishment and coercion’. She adds: ‘The carceral state is and must be constructed and reconstructed, brick by brick, cage by cage, count by count’ (p. 7). We proceed then by clarifying why and how the churning of prison infrastructure was instrumental for the NSW’s broader agenda of carceral expansion and consolidation.

Carcerality, coloniality, and ‘Better Prisons’ in NSW

The 1990s and 2000s were decades of significant growth in the number of people incarcerated in NSW. Between 30 June 1990 and 30 June 2009, prisoner numbers in NSW almost doubled, reaching 11,133 people. Much of this increase was due to a slow and steady rise in the number of people on remand, namely incarcerated but yet to be sentenced. This period of rising incarceration provided a pretext for an unprecedented wave of prison building in NSW. In a submission to the NSW parliament in November 2012, then Commissioner for Corrective Services NSW explained that, to manage the rising and fluctuating inmate population, the system had been expanded through measures such as reopening decommissioned facilities, expanding existing prisons, and constructing new ones (Severin, 2012: 2–3). The Commissioner argued that the result was a substantial net increase in the state’s carceral infrastructure with the number of prisons in NSW rising from 26 at the beginning of 1990 to 36 by 2010 (p. 8).

Between 2009 and 2011, the state saw a drop in its prison population. This decrease came as something of a surprise to policymakers and was mainly the result of a decrease in sentenced prisoners. Government criminologists explained the decrease as resulting from reduced offending rates for certain offenses such as assault, and in judges giving shorter sentences (Fitzgerald and Corben, 2012). Against the backdrop of prior carceral infrastructural expansion, by mid-2011 NSW had an excess of prison bedspace. Consequently, on 6 September 2011, the NSW Treasurer announced a plan to resolve the ‘surplus capacity’ by closing Berrima, Parramatta, and Kirkconnell Correctional Centres (Baird, 2011: 6). Subsequently, on June 29, 2012, the Government announced the additional downsizing of Grafton Correctional Centre due to the facility’s age (119 years old), high operational costs, and geographic location, which is being serviced by newer facilities.

Around this time, the NSW Government undertook significant bail reform, notably introducing the Bail Act 2013, designed to address the soaring remand population and the unwieldy complexity of existing laws, while re-emphasising the presumption of innocence (Bartels et al., 2018; Brown and Quilter, 2014). However, within just 5 weeks of its May 2014 operation, the 2013 Act faced a media storm following three high-profile bail decisions, leading to public and media outcry accusing the NSW Government of being ‘soft on crime’. This prompted an urgent review by former Attorney General John Hatzistergos, resulting in legislative amendments which effectively reversed key 2013 reforms. While the original 2013 reform aimed to reduce remand numbers, subsequent amendments were explicitly intended to make obtaining bail harder. Unsurprisingly, the number of incarcerated people in NSW began to rise once again, driven by an increase in the number of people held on remand.

While the legislative reform was a response to high-profile crimes, the absolute number of crimes occurring in NSW fell during this period. Survey data from the official statistical agency shows that for high prevalence offense types such as assault, break-in, motor-vehicle theft, and malicious property damage, the number of crimes occurring has trended down over the last 15 years (ABS Crime Victimisation, 2023–2024). Yet the number of incarcerated people has trended in the opposite direction.

Just 4 years after contracting the system through prison closures and downsizing, by 2015, in the context of a consolidated carceral infrastructure and the falling incidence of crime but rising imprisonment rates, the NSW Government launched a new wave of prison restructuring. Under the reformist banner of ‘Better Prisons’, the state dovetailed responses to both overcrowded prisons and a scathing audit report (Whitfield, 2016) that recommended the state better integrates principles of new public management into NSW’s carceral infrastructure (cf. Freiberg, 2005). A key element of Better Prisons was the $3.8 billion ‘Prison Bed Capacity Program’, announced in 2016. As the Minister told the NSW Legislative Council in September that year (2016):

Better Prisons is all about providing new jails. It is about creating a situation in New South Wales where we do not have pressures with custody. We have an allocated $6.2 billion budget over the next four years. That is a record budget. That includes $2.2 billion in capital which, over four years, will fund our prison bed capacity [. . .]. We need state-of-the-art facilities for the twenty-first century. We have to move beyond those Victorian era jails if we are serious about rehabilitation.

In all, 6100 additional ‘beds’ were to be delivered by May 2021 (Goodwin, 2019). The Better Prisons programme constituted a profound spatial and operational reconfiguration of the state’s carceral landscape, driven by dual imperatives of efficiency and a re-imagined, modernised ‘rehabilitation’. This staggering expenditure represented a monumental expansion of the carceral footprint, entrenching, and normalising mass incarceration under the guise of infrastructural improvement.

Figure 1 shows the growth of both the number of prisoners, and the ‘design capacity’ of prisons in NSW over a 29-year period. ³ It is clear that the wave of prison expansion which began under the banner of Better Prisons in 2016, had led by 2022–2023 once again to surplus carceral capacity. As in 2009, the number of ‘beds’ in 2023 comfortably outstripped the number of incarcerated people. The groundwork for either another round of carceral ‘expansionary-closure’, or further ‘tough on crime’ sentencing and bail reforms to fill that capacity, had been laid.

Figure 1.

Incarcerated adults and prison system design capacity, NSW, 1994/1995–2023/2024.

The Better Prisons programme exemplified an internal logic of prison expansion that produces future prison closure. Noteworthy was the deployment of short lifespan ‘rapid build dormitory style prisons’. Established within an 18-month timeframe and assembled from ‘preconstructed accommodation pods’ on existing prison grounds, these facilities underscore the temporal dynamics of carceral creative destruction (Goodwin, 2019: 13). While presented as providing ‘fit for purpose’ capacity, the over 1500 beds delivered under the rapid build programme were not considered suitable for long-term use—with an initially projected expected life span of 5–7 years. ⁴ Aimed at managing an overcrowding crisis, these facilities’ planned obsolescence is a stark example of the logic of the prison fix, an internally-produced guarantee that future prison building will still be needed. These short lifespan prison openings create a future in which prison decommissioning will quickly become a necessity, laying the groundwork for another wave of prison building.

That the reorganisation of prison infrastructure in NSW between 2011 and 2015 was expressed through a juxtaposition of old (inefficient, even inhumane) facilities versus new (modern, rehabilitative, effective) ones is critical. While the couching of prison reform as ‘improvement’ is expected, this trend is rarely understood through a settler colonial vernacular. In the case of the NSW prison churn, the closure of old prisons was presented as a move against and away from ‘the Victorian era’. However, these closures—as a first step in a series of subsequent decisions that included prison reopening, decommissioning, and sell off—were very much in line with the state’s long history of carceral colonialism. As Carlson (2023) shows in her analysis of the colonial history of Queensland, reforms cast as progress often entrench colonial power in new forms. The NSW prison churn thus represents not a break with the past but its continuation, as the system ‘shape-shifts’ (Gilmore, 2024) yet remains beastly as ever. We illustrate this dynamic through four case studies—Berrima, Parramatta, Kirkconnell, and Grafton—closed or downsized in 2011–2012.

Prison expansion in action: Four case studies

Below we take a closer look at some of the economic, political, and ideological aspects of the prison churn in NSW. The case studies illustrate how the consolidation and amplification of carceral state power were enabled by specific decisions to close or downsize prisons, as well as reopen them as temporary solutions for prison overcrowding. These decisions were carried out at the same time as the state massively invested in expanding its carceral infrastructure. Each case highlights a slightly different facet of the process, yet together they illustrate how crisis and prison churn are entangled at the local and regional levels with the economics of the spatial fix.

One: Berrima Gaol

Located 110 km south-west of Sydney, Berrima is a former colonial centre in the Southern Highlands, now rebranded around agritourism and boutique hospitality after the decline of agriculture. Built in 1839, Berrima Gaol served varied functions—from a WWI internment camp to a mid-20th-century labour camp—before being rebuilt in the 1940s and later reclassified as a medium- to minimum-security prison with community work programmes. In 2011 the NSW government closed the facility, citing its age and falling prisoner numbers as justifications. According to a report produced by the NSW Select Committee on the Closure of Downsizing of Corrective Services NSW Facilities (hereon, Select Committee), at the time of its closure in October 2011, Berrima held a total of 60 inmates, with a capacity of 75. These inmates were relocated to Dillwynia and Emu Plains Correctional Centres, and the 49 staff that were based at Berrima were redeployed or offered voluntary redundancies. The closure was accomplished with minimal cost to the government. Corrective Services NSW reported the closure incurred the following expenses: inmate escort costs of $203, infrastructure costs of $61,440, and on-going costs of $80,000 for the 2012/2013 financial year, until the site is handed over to the NSW Department of Primary Industries, Catchments and Lands (Select Committee on the Closure or Downsizing of Corrective Services NSW Facilities, 2013). In a submission to the Select Committee, Rosalind Dale, who was employed as a clerk in Berrima until its closure, described the value of the prison to staff, prisoners, and the regional community:

The Print Shop was a flourishing unit with many inmates learning new skills and it returned a good profit. [. . .] The inmates contribution to the local community was valuable to many organisations. They were actively engaged in the long term maintenance of Berrima by mowing public grounds for the local Council. The inmates participated in a program in conjunction with the National Trust maintaining the gardens at the local historic property, Harper’s Mansion. On an annual basis the inmates participated in planting over 10,000 tulip bulbs in parks in the Shire for Tulip Time, the major tourist attraction; the preparation of Moss Vale and Robertson showgrounds and steward duties during the shows and preparation of Bundanoon oval and assisting at the Scottish festival, Brigadoon and the Goulburn Rose Festival. (Dale, 2012)

Reflecting on the harm caused to the staff by the closure, Dale noted that its immediate impact on staff was devastating, with sacked workers facing the prospect of long-distance travel or uprooting their families from the region.

While the closure of Berrima was protested by the community and staff, it was carried out with minimal pushback. The long-term regional impact of the closure was not considered in the decision, nor was the wellbeing of incarcerated people who had to be displaced into new facilities, often located farther from family and support networks.

Less than 5 years later, the NSW government decided to reopen Berrima Gaol. The decision came following the period of sharp increase in prison population which led to severe overcrowding. Beyond presenting an ethical dilemma (in the form of prisoners’ health and wellbeing), overcrowding was regarded as a major risk to the safety of prison guards. As NSW Inspector of Custodial Services described in his 2015 report (Paget, 2015: 5):

[. . .] the institutionalisation of ‘doubling up’ as both a response to inmate population growth and as a design practice; excessive inmate movements around the estate; insufficient legacy staffing arrangement; budget pressures and saving targets, which have resulted in inmates being locked down at 2.30pm in the afternoons or on weekends; facilities which have not kept pace with the growth of the inmate population; and program provision to address offender behaviour which falls well short of requirements.

Where the state treats inmates in a way that denies them a modicum of dignity and humanity it should not be surprised if they respond accordingly, with individual acts of non-compliant behaviour escalating into collective disorder, such as riots.

While the Inspector’s report repeatedly emphasised that old, 19th prisons are structurally unfit for purpose, and that the overcrowding in these prisons elevated risks of harm, the NSW government responded to its self-made crisis of overcrowding by reopening Berrima, at least temporarily, to alleviate the pressure in the system. The ethical and financial costs of reopening Berrima were deemed justifiable in relation to risks associated with overcrowding, particularly considering the gaol’s additional value for the region.

The government’s decision to reopen the prison for 75 male minimum-security prisoners was welcomed by the local community. On their website, the Berrima Residents Association group stated it ‘expects to again work cooperatively with the prison authorities for the benefit of the village and the wider community in the Southern Highlands’ (Berrima Residents Association, 2016). Recalling Dale’s submission to the Select Committee, the positive response to Berrima’s reopening seems to reflect the fact that inmates were integrated into the local economy—willingly or not. Berrima reopened on 21 March 2016 and continued to operate on a similar capacity to its pre-closure days until it was finally decommissioned 5 years later, in March 2021. While we cannot discuss the final closure of Berrima in detailed here, it is noteworthy that the ultimate closure was met with considerable community pushback, particularly given the government’s decision to sell the prison to a private developer rather than grant it back to the community to operate and preserve (McCabe, 2025). In 2022, Blue Sox group purchased Berrima Gaol for $7 million dollars. It is now at advanced stages of redeveloping the site into a ‘hub of venues [. . .] designed to bring together the community and attract visitors to the region through intimate dining experiences, a wellness retreat, function rooms, gallery & retail, public bar and bespoke high-quality accommodation (Blue Sox Group, 2025)’.

Berrima’s story represents a classic case of the prison churn where chaotic policymaking tries to catch up to state manufactured crisis. But we can also see in this story the entanglements of the prison churn with the spatial fix, represented by three competing economic priorities: a local desire to extract cheap labour from inmates, the state’s attempt to minimise cost per bed by shutting down and then reopening old prisons, and a combined state and regional interest in selling and repurposing the prison to boost regional economic activity. We see below that a similar logic to the latter determined the fate of Parramatta Gaol.

Two: Parramatta Gaol

Like Berrima, Parramatta Gaol is a historic site in what is now a major commercial centre in Western Sydney. First built in timber in 1796—just 8 years after the invasion of NSW by the British—it was destroyed by fire, rebuilt in 1802, and quickly fell into disrepair. A new stone prison, completed in 1842, was expanded several times in the 19th century, becoming the colony’s second largest by 1897.

Parramatta was shut down in 2011, and like Berrima, its closure was not a costly operation. Done gradually over the course of a year, Parramatta reduced its capacity in September 2010 from 580 to 240 beds, with the remaining inmates transferred out of the facility in October 2011. Since the site was handed over to the State Property Authority in March 2012, it incurred no additional maintenance costs (Select Committee on the Closure or Downsizing of Corrective Services NSW Facilities, 2013: 22). At the time of its closure, Parramatta was the longest operating prison in the country. Given its location in Parramatta’s central business district, and its age and condition, the closure of Parramatta invoked little to no resistance from the community, and many of its 143 staff were successfully redeployed in other Sydney and outer Sydney prison facilities. In 2016 the ownership of the prison was transferred to a Local Aboriginal Land Council under the NSW Aboriginal Land Rights Act (1983). ⁵

In line with its emergency response to prison overcrowding, in 2016 the NSW government decided to reopen Parramatta Gaol. However, unlike Berrima, the suggestion to reopen Parramatta provoked a fierce public backlash. In March 2016, then Premier of NSW Mike Baird stepped into the microphone to reassure concerned property developers and the residents of Parramatta that re-opening the prison will not jeopardise the government’s plan for a major redevelopment of the area. Baird’s attempt to alleviate concerns was necessary given the government’s move (made in the preceding November) to rezone land around the prison to make way for a new housing estate that included 3000 homes and a village centre (Kembrey, 2016b). While the Premier insisted that the development plan will go ahead regardless of the prison’s reopening, the site’s developer, Urban Taskforce, criticised the government’s reopening proposal. As Urban Taskforce CEO, Chris Johnson, put it:

The NSW Government seems to now have two contradictory approaches for the renewal of North Parramatta. On one hand [. . .] new residential towers around the old gaol and the proposed light rail [that] will connect the precinct while on the other hand the government proposes to reopen the gaol which does not send a welcoming signal to incoming residents. (Urban Taskforce, 2016)

With investors convinced that Parramatta is set to become ‘a very desirable place to live and work in’, Johnson stressed that reopening the prison will undermine investors’ confidence and set any development plans back. Similarly, a property consultancy manager working in the area estimated that if reopened, the prison could bring property prices down by as much as 30%: ‘It would impact on the value that the owners of property would realise and [. . .] on the demand. Obviously, having a jail within close proximity to where you are living would make you potentially look elsewhere’. Western Sydney lobby groups slated the government’s proposal due to its potential damage to the area’s significant renewal efforts (Kembrey, 2016a).

Amidst mounting pressure by the development sector and local government, and fears that reopening Parramatta might jeopardise the massive investment already made by the state government in revamping and rebranding the area, the NSW Minister of Corrections finally withdrew the proposal to reopen the gaol. He explained that he couldn’t ‘reconcile . . . spending tens of millions of taxpayer dollars reopening the jail, with all the development happening in the area [. . .]’ (Adoranti, 2016). With Parramatta Gaol remaining shut, the renewal of the area is well underway, and its large-scale housing and transportation projects are near completion. The decision not to reopen the prison was directly linked to the government’s vision for the city of Parramatta. At the same time, this decision must be understood in the context of the state’s active construction and expansion of its penal infrastructure outside of Sydney (Whitbourn, 2016), and its $3.8 billion ‘Better Prisons’ programme discussed in the previous section. As Premier Baird clarified following the decision not to reopen Parramatta Gaol: ‘Over the long term we are building a very significant new facility up in Grafton [. . .]. We need to get to a position where there is surplus capacity in the [prison] system and that’s obviously what we’re working towards’ (Gerathy, 2016). Ironically, this is the same officeholder who, as NSW Treasurer, announced in 2011 the closure of prisons due to a surplus in the system. The ‘significant facility’ mentioned by Baird is the Clarence Correctional Centre, which opened in 2020 with a capacity of 1700 beds.

While similar considerations led to Parramatta’s closure in 2011 as to Berrima’s, Parramatta did not reopen. The context of urban capital accumulation was likely determinative. With the land being rezoned for residential and commercial development, and a booming speculative economy in the area, here the logic of the spatial fix kept Paramatta closed. Put simply, there was not much to be gained by refilling Parramatta with prisoners, but the potential losses were great.

Three: Kirkconnell Correctional Centre

Kirkconnell Correctional Centre is an outlier in this story. Built in 1961 as one of five Afforestation Camps offering inmates post-release skills, it later expanded to 250 beds. Unlike the 19th-century Berrima and Parramatta, Kirkconnell is mid-20th century and remains open today despite a temporary closure. Located 30 km east of Bathurst, the prison was shut in 2011 but kept ‘mothballed’ at a cost of $475,000 annually so it could reopen if needed (Select Committee on the Closure or Downsizing of Corrective Services NSW Facilities, 2013). Staff, inmates and the community were late to be informed about the closure, causing, as a former staff put it, personal, economic, and social harm (Craig, 2016).

The decision to reopen Kirkconnell in 2015 was bitterly criticised by state officials and local government representatives—not because they opposed its reopening, but since they thought it should have never been closed to begin with. The opposition Labor party’s justice spokesperson, Paul Lynch, referred to the decision to reopen Kirkconnell as an embarrassing admission that the government ‘got it wrong’ when it closed the facility: ‘Closing and reopening jails costs money, and for the Baird government to reverse course in a single term [. . .] shows its grossly incompetent planning’ (Hasham, 2015). The accumulated cost of keeping Kirkconnell functional following its initial closure was reportedly $1 million, and the decision to reopen it was accompanied by a significant $4 million investment by the NSW government, which covered extensive repair work and grounds upgrade, as well as a modest expansion of the prison to 260 beds.

Unlike the strong and swift pushback against the reopening of Parramatta, the reopening of Kirkconnell in July 2015 was enthusiastically welcomed by local elites. The state Member for Bathurst, Paul Toole, emphasised the anticipated local economic and demographic contribution of the reopened prison: ‘We’re seeing a lot of new trained corrective services officers coming into the area. [. . .] It is very exciting to see the job prospects that are being offered here as part of the re-opening’. In relation to labour opportunities for inmates, Toole explained:

A lot of community work’s taken place in the past, especially around those community halls, which are important facilities for the whole community that they can benefit from. This is important, because a lot of these areas are made up of volunteers, and if they can get some additional support [through inmate labour] this is a great way of actually working in collaboration and working together for a very worthy outcome. (ABC News, 2015)

While puzzling, Kirkconnell’s story is central to our argument. At first glance, it looks like policy failure, a costly decision made with little regard for long-term consequences. Yet the 2011 closure was tentative from the outset. By continuing to fund maintenance and security, Corrective Services effectively hedged against the uncertainty of a closure shaped by specific, reversible policies. This anticipatory logic of closure and reopening also appears in Grafton’s case, though there the community played a far greater role in shaping the outcomes.

Four: Grafton Gaol

Located 630 km north of Sydney in the Clarence Valley, Grafton Gaol began operating in 1893. Classified as a maximum-security prison in 1924, over the years it was variously criticised for harsh conditions and mistreatment of inmates. The announcement of Grafton’s downsizing came shortly after the closure of the previous three facilities. In June 2012 CSNSW Commissioner issued a media release stating Grafton was to be restructured as an ‘intake and transient centre for inmates in the northern part of the State and operate with a revised staffing structure’. At the time, the prison had an operating capacity of 243 inmates (sentenced and unsentenced), both medium and minimum security (Select Committee on the Closure or Downsizing of Corrective Services NSW Facilities, 2013: 23). Like other colonial-era prisons, the decision to downsize Grafton stated its age and condition, resulting in higher incarceration costs, as justification. As explained by Commissioner Severin, the ‘cost per inmate per day at Grafton was $203.24 in 2011-2012, [. . .] at Kempsey [. . .] $154.85 in 2011-2012, [and] at Cessnock [. . .] $142.86 in 2011-2012’. The Commissioner also stated that ‘geographically and logistically Grafton has not got a catchment area in relation to the prisoner population which would support the on-going management of the facility’ (in, Select Committee on the Closure or Downsizing of Corrective Services NSW Facilities, 2013: 23).

The decision to downsize Grafton was met with widespread criticism and organised protest, including a 5-day picket at the gaol by staff and community. In his evidence to the Select Committee—which was appointed in response to the backlash following the Grafton decision—Mayor of Clarence Valley Council, Richie Williamson, described the scene of July 2012 this way:

On 7 July . . . the community came together to rally at the gaol and that rally continued for a number of days and stayed in place until 12 July . . . [T]he Department was hell- bent on making [the downsizing] happen—and happen at all costs. On Sunday 8 July I witnessed what I will describe as a very dangerous situation when we saw a prison van move through a crowd of people, on the way allegedly causing some damage to a parked car. [. . .] With the vans entering the facility, my community stood in absolute silence—without any blockages whatsoever—to watch something that they had fought for over the previous days, simply disappear.

The downsizing of Grafton was costlier that the closure of Berrima and Parramatta. CSNSW reported the costs as follows: inmate escort costs of $116,055, infrastructure costs of $46,403, and on-going costs of $388,632 for the 2012/2013 financial year (Select Committee on the Closure or Downsizing of Corrective Services NSW Facilities, 2013: 23). Yet this move was projected to save considerable money for the government in the long run.

Much of the criticism levelled against the decision to downsize Grafton focused on the lack of sufficient research conducted in advance to evaluate the impact the decision would have on the community. As the Member for Clarence, Mr Chris Gulaptis, put it (in Select Committee on the Closure or Downsizing of Corrective Services NSW Facilities, 2013: 26):

On a decision of the magnitude of Grafton, I think in the first instance a rural communities impact statement needs to be made, and I believe that a decision like that should have gone through Cabinet. [. . .] Going through Cabinet would then enable all of the Ministers to become involved and then enabled me as a member of Parliament to have some involvement through that process . . . I felt that I was left completely out of the decision-making process.

Matt Bindley, State Chairperson of the Prison Officers Vocational Branch, Public Service Association and Senior Correctional Officer, used a harsher tone: ‘I do not understand how the Government could accept the ideology of closing any Correctional Centre let alone a rural Correctional Centre without any rural impact statements [. . .]. This was only ever going to leave a situation of bewilderment and confusion throughout the community’ (in Select Committee on the Closure or Downsizing of Corrective Services NSW Facilities, 2013: 26–27). Likewise, in their submission, the Clarence Valley Community Unions concluded: ‘The NSW Government should not go ahead with any decision to abolish rural job opportunities without first completing a rural impact statement on the effects of such a decision on the rural community’ (King, 2012: 5).

Despite widespread unrest, the decision to downsize Grafton proceeded, shrinking the inmate population from 243 to 64 and reducing the staff from 114 to only 38. Subsequently, the prison continued to decrease its capacity until it was finally closed in 2020, when the new Clarence Correctional Centre had opened. In its 95-page report of June 2013, the Select Committee made nine recommendations in relation to how decisions about prison closures, particularly regional ones, need to be taken. The recommendations included, among others, the inclusion of rural communities impact statements ahead of closure decisions (1), proper consultation with NSW Trade and Investment in relation to labour impact (2), and ‘That, should there be a need to build new correctional centres in the north of the State to accommodate the inmate population, Corrective Services NSW give consideration to planning a new facility in the Clarence Valley region’ (9). The appointment of the Select Committee and its recommendations suggest that the pushback against Grafton’s closure was to have a lasting impact on carceral planning in NSW. Indeed, recommendation 9 can be read as foreboding the construction of Australia’s first mega prison, Clarence Correctional Centre. Clarence, located only 17 km from Grafton, was built over the course of 4 years (2016–2020) at a cost of $700 million. Privately operated by Serco, the Centre is the largest in Australia and a flagship of the ‘Better Prisons’ programme.

Conclusion

This paper interrogated the role of prison closures and downsizing in carceral restructuring and the consolidation of carceral state power. Through a conjunctural analysis of the debates, dynamics, and processes surrounding four prisons across urban and regional NSW, and a broader look at trends of prison design capacity and incarceration rates over the past two decades, we showed that the closure of prisons, temporarily or permanently, has been used in NSW as a spatial fix for a range of economic and political crises. While at first glance appearing as the contraction of the carceral state (in 2011–2012), prison downsizing and closures served as a smokescreen that enabled the rapid expansion of the state’s prison infrastructure by thousands of beds over the course of less than a decade. The paper develops the concept of the ‘prison churn’ as the cyclical reworking of carceral assets, including closure, downsizing, mothballing, reopening, redevelopment, and decommissioning. Whereas the ‘prison fix’ captures the underlying logic of prison expansion—addressing the manifold ‘crises’ of racial capitalism and settler colonialism through investment in punitive infrastructure—‘prison churn’ sheds light on the temporal mechanisms through which this process is operationalised. We see that the churn can be fast moving, even frantic, in its seemingly contradictory workings across time and space.

In NSW, the prison churn allowed the state to variably transform bedspace capacity into fiscal capacity under the guise of infrastructural improvement, articulated specifically against a rundown Victorian-era infrastructure ‘no longer fit for purpose’. In a sense, this process is not entirely surprising. We know that prison reform is constitutive of the prison system itself (Davis, 2011; Foucault, 2012 [1975]). A resurgent of law-and-order politics, often in revanchist response to the modest gains of legal reformers, may generate renewed demand for incarceration. This dynamic is particularly acute in moments of social unrest or electoral competition, when punitive rhetoric serves as a proxy for state legitimacy. Simultaneously, prison construction functions as a spatial fix in rural and deindustrialised regions, where it is marketed as economic development. As Gilmore (2007) has shown, prisons are sited not where crime occurs, but where surplus land, labour, and political capacity can be mobilised for carceral ends.

Our analysis highlights the value of non-US prison research for understanding contemporary carceral restructuring. Much of the theoretical language used to analyse prison expansion has emerged from the United States, shaped by its specific history of racial capitalism and mass incarceration and its administrative order. American theory travels usefully, clarifying how prisons function as spatial fixes and as instruments of state crisis management rather than simple responses to crime. Nevertheless, reading the prison fix through NSW brings into focus processes that have received less attention in US-centred accounts, in this instance the role of infrastructural instability in sustaining and expanding carceral power. The Australian experience, marked by rapid prison growth and deep entanglements with settler colonial ideologies and governance, demonstrates the need for further regional, conjunctural analyses of prison expansion. Such work is essential not only to understand Australia’s unprecedented carceral build-up, but to strengthen and recalibrate the theoretical tools through which carceral restructuring is understood globally.