Rent regulation: unpacking the debates

Abstract

In the context of chronic problems with high housing costs, rent regulation has returned to the forefront of policy debate in several countries. This paper addresses three distinct questions related to rent regulation and the role of evidence. First, what are the drivers of policy change on rent regulation and what role does evidence play in shaping change? Second, what is the nature of the evidence base on rent regulation and the key messages that emerge from it? Third, how is this evidence base transmitted into policy debate? We take the example of the recent UK policy debate to examine this issue. The paper discusses the case of current policy development in Scotland to reflect upon policy drivers and the role of evidence. The paper combines insights from a recent evidence review and a decade-long policy ethnography. Not only does research indicate that evidence has played a limited role as a driver for policy change on rent regulation but the nature of the evidence base is such that there are limits on the guidance it can offer and the extent to which policy can in principle be rooted in evidence. The case of Scotland illustrates the forces at play in shaping rent regulation policy.

KEYWORDS: :

• Evidence
• policymaking
• rent controls
• Scotland
• economics

Introduction

While there has been a broad trend in housing policy towards liberalisation of private rental markets over the last forty years, rent regulation has remained a feature of many national housing policy systems (Kettunen & Ruonavaara, 2021). In a context where lack of housing affordability continues to be a global concern (Wetzstein, 2017; Gibb & Hayton, 2017), the reintroduction or strengthening of rent regulation has been on policy agendas at national, state or local level in countries including the Republic of Ireland, Germany and the USA.

Yet rent regulation is more popular in policy and tenant advocacy circles than it is from an investor or landlord perspective, or among economists and market commentators. Sharp differences between stakeholders over the merits of the policy make it a particularly contentious aspect of housing policy (eg. McPherson, 2004; Grey, 2018). It is a debate that occurs at the intersection of the technical, the normative and the fiercely political.

This paper focuses on three distinct but inter-related questions relating to policy change and the role of evidence. First, what are the drivers of policy change on rent regulation and, in particular, what role has the evidence base played in shaping change? Second, what key messages emerge from the current literature on rent regulation? Does the literature offer sufficient clarity in its messages that it could offer guidance to rent regulation policy? Our aim here is not to discuss findings in detail, rather we focus on the nature of the literature as a body of evidence. Third, we consider how the evidence base has been transmitted into the policy debate. This issue touches upon the scope for policy learning and policy transfer and raises key epistemological questions. The paper seeks to contribute to our understanding of how rent regulation policy is shaped, in the light of aspirations for evidence-based policy making. The core questions we address are of broad relevance, but to illustrate the issues we discuss recent policy debates and developments regarding rent regulation in the UK and then consider Scotland more specifically as an illustrative case study. The paper draws on a review of the recent academic literature on rent regulation in economics and the social sciences more broadly (Gibb et al., 2022) and a small-scale policy ethnography (Rhodes, 2011).

In this paper we use the term “rent regulation” to refer generically to policies aiming to influence the levels or rates of increase of private rents. Such regulatory policies in practice take several forms and pursue different objectives (Lind, 2001). A frequently employed categorisation distinguishes between first-, second- and third-generation rent controls (Arnott, 1995). First-generation controls represent nominal rent freezes; second-generation controls are regulations that allow rent increases for specified reasons and do not allow decontrol between tenancies; third-generation controls represent intra-tenancy rent regulation with inter-tenancy decontrol. Where rent controls have been recently (re)introduced they are typically third-generation rent controls. While the term ‘rent control’ is still applied to these diverse arrangements, when allowable increases under second- or third-generation rent controls are tied to inflation or maximum percentage increases they can also be referred to as ‘rent stabilisation’ policies rather than rent control.

We are principally interested in rent regulation, which is the focus of much of the relevant economics literature. However, in practice, approaches to rent regulation are embedded in a broader institutional and regulatory context, encompassing regulation of property standards, security of tenure, acceptable management standards, and tenant rights and obligations (Whitehead et al., 2012). Moreover, in practice, when rent regulation changes it is often accompanied by changes across other dimensions of regulation. Isolating the effect of regulating rents per se is challenging. We return to this point when we discuss the evidence base.

The paper is structured into a further seven parts. It first briefly summarises key messages from the literature regarding policy change and the role evidence can play in it. It then provides an overview of our review method. The paper next discusses the research evidence regarding the drivers of policy change on rent regulation. The following two sections consider the nature of the evidence base and how it has been transmitted into policy debate in the UK. The penultimate section turns to consider the case of current developments regarding rent regulation in Scotland. The final section concludes.

The role of evidence in policy change

The potential sources of policy stability and change are diverse and complex. John (2013) grouped them into five broad categories: agency/rational choice; groups and networks; institutions; ideas; and exogenous factors. The most prominent theories of the policy process synthesise factors from two or more of these categories (Weible & Sabatier, 2018). Different theories allow different scope for evidence to be influential, and to be so in different ways. Those viewing policy as driven by material interests or electoral calculation might ascribe a limited independent role to evidence, whereas those emphasising ideas or rationality open up greater scope for evidence to influence policy directions (Cairney, 2016).

The research utilisation literature identifies several models for the relationship between evidence and policy (eg. Best & Holmes, 2010). The stream of argument focussing more directly on the political uses of evidence (following Weiss, 1979) views evidence as being selected ex post to rationalise policies pursued for other reasons: an example of so-called “policy-based evidence” (Strassheim & Kettunen, 2014). Boswell (2009) highlights the way in which evidence is invoked selectively not to shape policy but to legitimate and substantiate existing policy positions.

Simplistic invocations of evidence-based policy assume that “the evidence” is comprehensive in its coverage, unequivocal in its messages, and clear in its implications. Yet this downplays the extent to which there are evidence gaps, different parts of a literature address different concerns, and/or point to different conclusions (Boaz et al., 2019; Lees, 2012). These considerations intersect with key concerns in the policy transfer literature: when evidence is reporting policy effects in a ‘source’ setting to what extent do those effects depend on contextual factors? If context is key then it means policy mobility to a distinct ‘recipient’ setting will have uncertain or unpredictable consequences (Soaita et al., 2021). Furthermore, where data require a theoretical framing to allow interpretation we can find ourselves facing “duelling certitudes” (Manski, 2013): those adopting different theoretical positions derive equally firm, but divergent, conclusions from the same evidence.

The research utilisation debate captures important components of the evidence and policy relationship. But, given the diverse factors shaping policy, research typically has only limited direct influence over policy (Boaz et al., 2019). This is why aspirations were moderated during the 2000s towards “evidence-influenced” or “evidence-informed” policy. Other factors unavoidably influence decision making alongside “the evidence” (Cairney, 2016; Pawson & Hulse, 2011). A more nuanced perspective recognises biases in evidence uptake (Stevens, 2007) and a more expansive understanding of what constitutes “evidence”. Within the academic community ‘evidence’ means systematically-collected research conducted by credible research institutes, but for policymakers valid ‘evidence’ can be considerably more diverse. Mulgan (2005, p. 219) noted nine policy-relevant sources of knowledge: statistical; policy evaluation (eg what has (not) worked previously); scientific; professional; public opinion; practitioner insights; political; economic; intelligence (eg regarding hostile powers or terrorists). Several of these intersect with “evidence” understood as research, but others represent knowledge of a different type. Cairney (2019) argues that appreciating the dynamics of the policy process means recognising that binary understandings of “evidence-based policy” versus “policy-based evidence” ignore a range of intermediate positions: decisionmakers value evidence but face imperatives to act that mean the purity of “evidence-based policy” cannot be realised.

When considering evidence and the policy process we also need to attend to how, and by whom, evidence is packaged and transmitted into the policy system. There is a key brokerage role in turning academic research into policy analysis digestible to the policy system. Policy advisory systems differ across countries and policy sectors (Craft & Howlett, 2013; Brans et al., 2017). In addition to analysis generated inside government, other actors – business associations, think tanks, charities and voluntary sector organisations, university-linked policy institutes – offer perspectives on current and future policy directions that are ostensibly evidence-based. Processes of evidence synthesis and representation into the policy process can – intentionally or unintentionally – shape the policy debate. In the absence of rigorous systematic reviewing, the scope for drawing different conclusions from the same literature is considerable.

Methodology

We attempted to provide the first large-scale review of the international evidence on rent regulation since Turner and Malpezzi (2003) while expanding beyond their focus on the economics literature to include research in the social sciences more broadly, including from the fields of political economy, social movements and geography.

Informed by commonly recognised (Dewilde, 2017) but still contested, country typologies, we scoped this review to selected countries belonging to three housing regimes, referred to as Liberal (AUS, CAN, GBR, USA), Nordic (DNK, NDL, NOR, SWE) and Continental (AUT, CHE, FRA, DEU). While in principle it would be possible to seek to encompass further, distinctive housing institutional arrangements (such as those of eg. India, Brasil or China), the countries selected offer both substantial diversity and areas of similarity in their historic and current rent-control arrangements and therefore offer grounds for scholarly reflection (Crook & Kemp, 2014). To identifying the relevant literature, we searched two large databases (SCOPUS and Web of Science) systematically for ‘rent control’, ‘rent stabili(s/z)ation’ and ‘rent regulation’ in titles, abstracts and keywords for literature published since 1990, limited to material published in English. We retrieved 402 unique records (of these 281 published since 2000). After checking for country, methodological and thematic fit, 185 records were maintained (116 published since 2000). Given available resources we decided to focus on the most relevant records published since 2000: we avoided duplicating Turner and Malpezzi’s work (2003). Each author independently scored these records for relevance on a scale of 1 to 3; all 70 references that scored 8 or 9 were agreed for reviewing (for more detail see Gibb et al., 2022). As is common practice (Soaita et al., 2020), this sample was supplemented by additional, expert-recommended studies (12 studies from 2003-2019 were added; eight on international comparisons, three on Britain and one on Ireland). We were aiming for comprehensive coverage of the relevant English-language literature on the sampled countries. Overall, 82 publications were reviewed, of which 18 were cross-country analyses.

Figure 1 presents two key features of the reviewed literature. The world map indicates the geographical focus of the 82 studies, with USA being dominant, followed at some distance by Germany, Britain and Sweden. There is thus over-representation of countries classified as within a Liberal regime versus those within the Nordic or Continental regimes. Within each regime, their recognised ‘prototype’ is also over-represented, i.e. USA (Liberal), Sweden (Nordic) and Germany (Continental). The left panel of Figure 1 indicates that most publications used quantitative methodologies (68%).

Figure 1. The reviewed literature: geographical representation and methods employed.

Notes:

• The left panel illustrates the dominance of quantitative studies (68% of total). Moreover, only 11 of the 26 qualitative studies used people-centred methods (e.g. case studies, interviews, ethnographies) with only a few interviewing tenants or their representatives.

• The right panel illustrates the different geographies researched: a national focus is dominant in Europe but less so in USA where many studies focus on cities. Taking into account that 18 references are cross-country analyses and that 13 advanced theoretical-only models, countries’ representation adds up as follows: USA = 38; DEU = 22; GBR = 18; SWE = 15; DNK = 9; FRA = 9; NDL = 9; NOR = 7; CAN = 5; AUS = 4; CHE = 3; AUT = 3.

To examine the case of current policy development in Scotland a policy ethnographic approach was adopted (Rhodes, 2011). This involved the collection of policy documents issued over the last decade; observation of Parliamentary and committee proceedings attended by the second author since 2016; and discussions with more than a dozen policymakers, national and local government officials, and other stakeholders in the policy community over the period of interest (2016-2021). To preserve confidentiality, our discussion in the penultimate section of the paper will focus on the broad contours of policy development. It makes no reference to individuals or specific interactions. As in any ethnographic work, the interpretation of these rich data belongs to the authors alone and does not necessarily reflect the views of any individual stakeholder or institution.

The drivers of rent regulation policy

Rent regulation policy has its origins in popular protest: for example, the 1915 Glasgow rent strikes were transformative for housing policy in the UK (Grey, 2018). But what does the recent social scientific literature say about the drivers of rent regulation policy and the role that evidence plays in those policy decisions?

Influenced by specific epistemological orientations, the literature offers two main, distinctive perspectives: one framed in terms of regime theory, the other within relational theories of practice and struggle. The former can help understand the importance of the broad institutional context within which the private rented sector is embedded as well as how change did and can occur (eg Kemp & Kofner, 2010). Historically, all our sampled countries have moved from first- to second-generation rent control or full deregulation (Voldman, 2013), with Liberal regimes tending to have moved more decisively in this direction. Considered comparatively, Sweden and Germany nonetheless stand apart because of their continuing strong rent regulation regimes (Martin et al., 2018). Change is also currently under way: some deregulated rental regimes are moving towards rent stabilisation (e.g. Ireland and Scotland); a few rent-regulated regimes are further strengthening their provisions (e.g. Germany); while in the US locally-regulated rental markets are diverging as some move towards deregulation and others reintroduce regulation. Although regime theory is grounded in the idea of political struggle, in the analyses we reviewed the politics of policy stability and change that underpin distinctive regimes remained relatively obscure.

The second perspective, inspired by relational theories of practice, offers finely-grained analyses. Whether coming from socio-legal studies (eg McPherson, 2004) or human geography (eg Teresa, 2016), these analyses evidence that changes in rent regulation have always been contested politically. These studies document the ways in which the political battle is fought in diverse venues (e.g. streets, closed consultations, TV screens) and among diverse social actors (e.g. legislatures, political parties, interest groups, lobbyists, think tanks, citizenry). Detailed analysis shows the primacy of discourse over evidence in the staging of election campaigns and public consultations, the formation of governmental coalitions, and the production of public opinion. This literature demonstrates that it is ideology - understood broadly as dominant beliefs endorsed by key policy actors or society - which constructs a private tenancy as a home or as a commodity and it is here that the parameters of the debate within a particular society, and the boundaries of regulatory space, are shaped (eg Bryant, 2004). This conflict can operate across different levels of governance as national governments wrestle with lower tiers of government for constitutional primacy over rent regulation (Fondation Abbe Pierre & Feantsa, 2021).

Overall, these studies indicate that evidence, understood as systematic research, has not played a prominent role in shaping policy. Policy decisions over the desirability or otherwise of rent regulation are not best understood as a matter of expert knowledge or econometric analyses but as a product of contextual factors and political struggle occurring both within and beyond the structures of representative democracy (eg McPherson, 2004; Voldman, 2013).

The evidence base on rent regulation

Whilst the social science literature indicates that the evidence base has not played a prominent role in shaping rent regulation policy, this might be because policy in this field is unavoidably driven by the contentious politics of conflicting material interests, but it could also be a product of the nature of the evidence base. To what extent does the evidence deliver clear messages about the effects of rent regulation that could, in principle, provide guidance to policy?

The evaluation of rent regulation within economics demonstrates a peculiar characteristic. There is persistent belief among economists in general that “rent control” is an ineffective policy mechanism that has an unambiguously deleterious effect on the housing market. There is therefore limited support for it. Yet, the range of opinions among specialist housing economists is markedly greater. There are surely those who hold similarly negative views, but most recognise that speaking of “rent control” lacks nuance: it misses important differences in regulatory design. First-, second- and third-generation controls will interact with the institutions of specific housing markets to produce distinctive outcomes. The latter perspective acknowledges that well-designed rent regulations can shape the way in which housing markets operate without having substantial negative impacts (Whitehead & Williams, 2018). Our review focussed upon those engaging directly with private rented sector issues, which represents the more specialist housing economics literature.

A rich post-war theoretical literature on rent regulation continued to develop until the mid-2000s (much of it reviewed by Turner & Malpezzi, 2003). This literature offers a variety of models exploring the impacts of rent regulations of different designs on a range of variables (eg property quality, new supply, prices in related markets) in housing markets with different assumed characteristics. The majority of models start from the assumption that housing markets can be characterised as perfectly competitive (eg Glaeser & Luttmer, 2003), while some assume that landlords engage in monopolistic competition or hold monopoly power (e.g. Basu & Emerson, 2003). Once these market structure assumptions are coupled with the specification for a rent regulation mechanism, conclusions can be derived. And substantially different conclusions can follow from different assumptions. For example, Arnott’s (1995) key contribution - that second-generation rent control can be relatively benign - flows, in large part, from respecifying housing market structure as less than perfectly competitive. The theoretical literature demonstrates that rent control can have a negative, positive or neutral effect on housing and related markets, depending on the assumptions you feed into the model. This characteristic of the literature led Lind (2007) to be sceptical regarding the potential of theoretical exploration in isolation from empirical work to advance our knowledge of the impacts of rent regulation. This literature exemplifies McCloskey’s critique of theoretical debates in economics that are no more than a “search through the hyperspace of conceivable assumptions” (McCloskey, 1991, p10). Empirical research is needed not only to examine regulatory impacts but also to shed light on the plausibility of underlying modelling assumptions to assist in model selection.

Econometric evidence on the impacts of rent regulation is relatively plentiful (for earlier reviews see Turner & Malpezzi, 2003; Jenkins, 2009). In contrast to these earlier reviews, rather than focus on substantive findings, here we reflect on the characteristics of the evidence base as a body of knowledge. We draw five main points from the evidence base we reviewed.

First, the empirical evidence is drawn heavily from the relatively small number of North American jurisdictions where rent regulation has operated, particularly New York, New Jersey, Massachusetts and Northern California (eg Ambrosius et al., 2015; David et al., 2014; Diamond et al., 2019; Glaeser & Luttmer, 2003; Heskin et al., 2000; Sims, 2007, 2011). There are also a small number of recent European studies (e.g. Fitzenberger & Fuchs, 2017; Lind & Hellström, 2006; Mense et al., 2019; Oust, 2018). The evidence indicates the importance of context, history and institutions and how they interact with any given form of rent regulation. The literature has also recently highlighted supply-side diversity – the distinction between corporate landlords and small portfolio landlords – in understanding responses to rent regulation (Diamond et al., 2019). The impact of rent regulation is not seen as universally negative.

Second, over the last 20 years empirical work has been more closely concerned with analysing second- and third-generation controls which have also typically entailed changing tenancy conditions. Econometric models are now more nuanced and context-specific. The effects of rent regulation can be difficult to disentangle from the effects of the non-price elements of the regulatory ecosystem, such as moving to longer tenancies or effective enforcement. In practice, components of a regulatory regime often change simultaneously. Holistic evaluation of such regime change is possible but isolating the impact of rent regulation can be methodologically challenging. There is consequently a risk of misreading the magnitude of its effects. Similarly, the impact of a rent regulation regime is the net effect of policy design and implementation, but direct consideration of regulations as implemented – addressing, for example, compliance and enforcement – is limited. Policies of the same design may have differential impacts in different places because of implementation differences. Earlier iterations of rent control in the UK, for example, highlighted non-compliance and gaps in implementation (Kemp, 1990; see also Byrne & McArdle, 2022). Evaluations which fail to account for these issues – or which uncritically mix the evaluation of imperfectly implemented existing policy with the benefits of an alternative that assumes perfect implementation – risk drawing conclusions of limited utility. Furthermore, the extent to which quantitative evaluation seeks to capture more elusive indirect effects – such as the impact of policy risk or the prospect of future (de)regulation on the current behaviour of market actors – is similarly limited.

Third, the battery of econometric tests – including pseudo experimental approaches - has grown substantially in the last 20 years to focus more on welfare effects, resource misallocation, impacts on uncontrolled adjacent markets and tenures, mobility and labour market effects, and neighbourhood impacts such as segregation (e.g. Glaeser & Luttmer, 2003; Skak & Bloze, 2013). These are valuable analytical developments but again their conclusions must be understood in the light of the empirical, regulatory and market contexts from which the data are drawn. Equally importantly, empirical studies typically focus on one or a limited number of policy impacts: holistic evaluation is substantially more difficult and hence largely absent. Furthermore, the specification of the (policy-off) counterfactual is integral to the assessment of policy impact and yet counterfactuals - e.g. outcomes in a competitive market with low-income housing vouchers – are rarely scrutinised or assessed for plausibility.

Fourth, we note a contrast between, on the one hand, increasingly sophisticated econometric modelling and, on the other, continued use of quite strong assumptions in order to make comparisons with control groups (e.g. using imputed rather than market rents), or about specific behavioural responses such as landlord substantive rationality, or about competitive market structures. Work attempting to make these foundational assumptions explicit and empirically-grounded remains rare. While the empirical literature has evolved considerably since the Turner and Malpezzi (2003) review, one of their key observations remains germane. They noted that analysis grounded in non-competitive modelling settings was rare. That absence of engagement with market structure remains.

Fifth, behind the story of rent regulation lies conventional welfare economics. This underlines the need to assess the extent to which real rental markets are as competitive as the theory assumes and, if they are not, the extent of market failures. Conventional welfare analysis argues that cash transfers are superior to in-kind subsidies or taxes on supply (i.e. rent regulation): what are the welfare considerations of plausible or credible (cash transfer) benefit systems that might be the means-tested alternative to rent controls? If these are no more than a theoretical counterfactual, they should have less weight in the argument. Relatedly, what impacts do possible rental market interventions have if we are dealing, in reality, with a second-best interdependent housing system where one shock or policy intervention has wider repercussions? It is not clear what this means for the relative merits of different models of rent regulation or the merits of rent regulation compared with other feasible, but less than perfect, interventions.

In summary, while these theoretical and empirical results point to a range of negative consequences from rent regulation, and the weight of argument lies on that side, that conclusion is by no means universal. Equally importantly, the evidence rests on a relatively narrow base in terms of housing market context, institutional context, policy design, and theoretical framing. Some conclusions can be generalised with a degree of confidence – for example, it is highly probable that crude first-generation rent controls will have negative consequences – but the number of such conclusions is relatively constrained. Advances in econometric technique have broadened the scope of analysis and allowed a more diverse array of impacts to be investigated, but holistic impact evaluations are lacking. The assumed counterfactual is crucial to the evaluation of policy impacts, and is strongly influenced by both theoretical framing and assumptions about policy feasibility, but receives less attention than it deserves.

Turner and Malpezzi (2003, p46) observed with respect to rent control: “if you are studying it, and you are not at least a little bit confused, you are probably not yet thinking clearly”. We would not characterise the current situation as a source of confusion. Rather it is one that, firstly, demands avoiding simplistic general conclusions about policy effects and, secondly, necessitates recognising that key characteristics of the evidence base mean that established conclusions, while providing valuable insights, can as yet offer us only a partial understanding of the net effects of rent regulation policy.

Transmitting evidence into the policy debate

In this section we examine how this evidence base has been transmitted into a specific policy debate by reviewing key contributions to the UK grey literature over the last decade. Private rented sector policy has evolved in the direction of greater regulation over this period. In Scotland this includes a shift towards the reregulation of rents and tenancy length (discussed further in the penultimate section). However, while increased rent regulation was included in the opposition Labour party’s manifesto for the most recent UK General Election, there is yet to be policy change in England and Wales. The increased policy salience of rent regulation within the UK and its devolved administrations has been accompanied by an increased volume of grey literature on the topic. Direct rent regulation has been almost entirely absent from the UK private rented sector for thirty years and, as a consequence, arguments about the virtues of the policy do not have recent local evidence to draw on – hence, arguments by necessity draw on evidence generated elsewhere.

In a report for the Residential Landlords’ Association (RLA), a key business association, Ball (2013) offers an economic critique of the case for long term rental contracts buttressed by inflation-only rent increases. He argues the policy would be detrimental to the expansion of rental market supply, most affect disadvantaged households, and increase risks for landlords. Moreover, Ball rejects claims about the benign effects of controls in continental Europe, where he argues investment is often weak and landlords are unable and unwilling to respond to rising demand because of rent caps. Bourne (2014), writing for the Institute of Economic Affairs, the free market think tank, draws on Ball (2013) and Jenkins (2009) to present a similar argument, framed as a critique of Labour proposals for third-generation rent controls. He argues that the apparent success of this type of system in Germany is not relevant because the institutions of the broader German housing system are radically different from those of the UK. Clay and Smith (2019), also writing for the RLA, present an argument that aligns with the conventional economics perspective: rent controls are bad for markets and communities and for the welfare of landlords and tenants. The authors point to a range of negative consequences including black market operations and ineffectiveness because, despite controls, rents continue to grow quicker than wages. Evidence is marshalled from North America (e.g. Jenkins, 2009) and from Europe, including some more recent evidence on Berlin’s rent brake.

In contrast, Scanlon and Whitehead (2014) address the case for rent stabilisation policies and longer tenancies in a publication for an academic research institute. They note that while schemes vary hugely across Europe, they are all very different from first generation rent controls and tend to provide both landlord and tenant with greater certainty. The authors argue this is useful for landlords: for example, institutional investors benefit from reduced tenant turnover and index-linked rent increases, although these benefits will not flow so easily to small landlords. Scanlon and Whitehead argue that it is not price regulation alone but a whole range of non-price regulatory factors combined with price stabilisation that determine the effectiveness of the overall rental market system.

Whitehead and Williams (2018, 2019), again writing for the RLA, highlight that models of rent regulation present a series of trade-offs for policy makers and market actors. The authors argue there is a ‘wide spectrum’ of approaches operating in different market and institutional contexts, with the UK sitting at the least regulated end of the spectrum and the Netherlands at most regulated end. They make a case for indefinite tenancies (as in Scotland) with a ‘degree of stabilisation’ and strengthened regulatory enforcement. In their second report (Whitehead & Williams, 2019), drawing on evidence from several countries in Europe and beyond, they ask what a ‘good’ tenancy might look like from an English perspective. They develop their 2018 conclusions by arguing further that the private rented sector is in practice highly varied and that reforms need time and resources to bed in. Both incentives and enforcement will be required to bring landlords along. Similarly, Wheatley et al. (2019), writing for a progressive think tank, consider how best to design a rent control system for London. They note some of the well-rehearsed risks and negative consequences critics associate with rent regulation policies, but argue these can be mitigated by careful design. They draw on international experience of rent regulation systems to try to identify the characteristics of a system that would work in the London context. Clarke et al. (2015), in a publication from an academic research institute, develop the theme of the importance of regulatory design. They simulate the effects of six different rent regulation systems. They argue that the impacts of different approaches cannot be treated simply as good or bad. Their provisional analysis indicates that a hard rent control would have a substantial negative impact on the sector but that, to differing degrees, other softer regulations would not.

The Rent Pressure Zone regime introduced in Scotland in 2017 is the focus of an independent assessment by Robertson and Young (2018), conducted for a major housing charity. They also draw on European evidence on different rent regulation regimes. They conclude that where “harder” rent controls have been introduced in urban hot spots landlords have not always complied nor authorities enforced. Where demand pressure is less intense rent regulation seems to work relatively effectively. However, they argue that the main issue is having a system that generates sufficient aggregate housing supply across all forms of housing to ensure that market constraints – rather than regulatory constraints - continue to exert downward pressure on rising rents.

Contemporary debates regarding rent controls are further synthesised by House of Commons library briefings (Wilson, 2017; Wilson & Barton, 2019). These publications aim to inform elected representatives in Westminster by providing even-handed summaries on key policy issues. They aim to localise the debates by embedding them in a discussion of the characteristics of the English private rented sector (Rugg & Rhodes, 2018). The papers draw on the syntheses offered by others (Turner & Malpezzi, 2003; Ball, 2013; Whitehead & Williams, 2018) but also turn occasionally to original sources (eg Diamond et al., 2019) in order to illustrate the arguments for and against rent regulation.

This brief overview illustrates the way in which authors have drawn on the same evidence base to arrive at sometimes starkly different conclusions. None of these reports attempts a systematic review and hence their engagement with the literature is selective. Nonetheless, the reading of the evidence base across the contributions does not necessarily differ substantially. For example, there is a level of agreement about the nature of the risks associated with rent regulation. The differences arise as a result of the relative prominence given to different elements of the literature; the assessment of the size and importance of the negative effects/risks in practice; and the implications drawn from the literature for future policy development. In some cases one can map the arguments promoted by an organisation on to the interests it represents, but that mapping is not perfect. The grey literature is not simply an exercise in legitimating – in Boswell’s (2009) terms – pre-existing positions. Most notably, the RLA, the business association, has published reports whose authors occupy somewhat different policy positions, rather than inevitably leaning towards the landlord interest.

The way in which arguments against rent regulation invoke previous reviews is notable. Jenkins (2009) is a more common point of reference than Turner and Malpezzi (2003). Jenkins’ conclusions are more unfavourable towards rent regulation so this is perhaps not surprising. However, a key point is that Jenkins’ review is explicitly and intentionally circumscribed in its coverage. It examines the US literature. To read directly across from its conclusions to the likely implications of rent regulation in the UK implies that contextual differences are not significant. Simultaneously, reasons are offered for disregarding apparently successful systems elsewhere because of contextual differences. This stands in contrast to those who point to examples of successful, or at least sustainable, systems of rent regulation.

While some such apparently successful systems are in North America, they are more likely to be operating in the Nordic or Continental countries in our sample. Looking to evidence from these systems carries the symmetrical risk: the assumption that if a system works elsewhere it could work in the UK. However, authors looking to successful systems tend to demonstrate an acute awareness that simply transposing a policy from elsewhere is unlikely to succeed, and that policy effectiveness rests on the broader institutional framework or regime (e.g. Wheatley et al., 2019). Hence, system design needs to be tailored in the light of local characteristics.

We can identify three core issues that divide positions in the debate. Firstly, can regulatory schemes be tailored to deliver an effective rent regulation policy within the institutional framework of the recipient country? Second, how important in practice are the negative effects/risks associated with rent regulation and are they sufficient to compromise the policy’s main objectives? Third, is it possible to mitigate the acknowledged risks of rent regulation through good policy design?

Our conclusion is that these core questions cannot be fully settled by appeal to the evidence base. Where there is evidence of apparently successful regulatory policy or of design strategies that minimise and mitigate risks we have to make further strong assumptions regarding the generalisability of those research findings if they are to be used to justify policy transfer to a new context. This is not an issue that can be resolved by filling gaps in the evidence base. Whether policy will succeed in a ‘recipient’ country is a question subject to irreducible uncertainty until the transfer happens in practice; and indeed dependent on how ‘success’ is define prior to and after change.

Direct application of conclusions from the evidence base to policy across contexts needs to be treated with great caution. Empirical evidence pointing to the conclusion that rent regulation has a negative – or a neutral or a positive – effect upon a housing market needs to be engaged with critically. While the evidence for that particular regulatory mechanism implemented in that place at that time might be clear, to what extent can conclusions be applied to other housing market contexts at other times? Rather than direct application, the evidence base can better be thought of as a sensitising device that highlights the need for policymakers to have a deep understanding of the contingencies – in market conditions, spatial scale, institutions, policy legacies and design – if local rent regulation policy is to be effective. It can also sensitise policymakers to practical questions of policy design and the areas in which problems can arise, but it cannot necessarily supply clear answers regarding what will work here and now. In this respect, pure evidence-based policy, even if desirable, is an unachievable ideal. Yet, evidence-informed policy – if there is space and time for it - is a feasible prospect.

A contemporary case study – rental market regulation in Scotland

Over the last decade Scotland has significantly departed from the rest of the UK in its approach to regulating the private rented sector. This has arisen from a long-term inclusive approach to a rental market strategy (published in 2013) that sought to improve standards and manage growth, seeking to balance the interests of tenant and landlord, while improving the position of tenants. Reform has primarily constituted a succession of non-price regulatory reforms (see Gibb et al., 2019). This culminated in the Private Housing (Tenancies) (Scotland) Act 2016 which created a new tenancy, ending no fault evictions (after an initial cooling off period) and, apart from 18 specific grounds for repossession, has arguably created a tenancy of indeterminant length. The new tenancy was accompanied by a form of third-generation rent control – albeit allowing generous uprating – that was a variant of Ireland’s local rent pressure zones (RPZs).

Local authorities in Scotland can create RPZs by applying for their introduction (at any spatial scale within authority boundaries) provided they have evidence of sustained unaffordable rent and households facing material financial hardship. If approved by the Scottish Government, the authority can then set a real terms cap on subsequent rent increases. Boyle (2020) identifies several years of sustained real increases in private rents in both Glasgow and Edinburgh, while Dick et al. (2020) suggest a wider range of local authorities are experiencing sustained real rent increases. However, the evidentiary hurdles to justifying designation have been sufficiently daunting that no local authority has yet applied for an RPZ (Scottish Housing News, 2018). There has been widespread criticism of RPZ policy for being ineffective (eg. Robertson & Young, 2018; Wightman, 2020; Dick et al., 2020). Wightman (2020) reports on Freedom of Information requests to the Scottish Government and all 32 local authorities: he found variable engagement with setting up an RPZ, but there are examples of several councils actively examining the possibility and some discussing it with the Scottish Government. Yet no council has actually introduced such a scheme.

One response to the perceived shortcomings of RPZs was the 2020 Fair Rents (Scotland) Bill, led by a private member of the Scottish Parliament (Pauline McNeill, Labour). This had reached the committee scrutiny stage when it was timed out by the end of the Parliamentary term in March 2021. It proposed replacing the RPZ system with a national CPI + 1% limitation on rent increases (a third-generation rent control). The Bill would also have made all contracted rents public information and allowed rent officers to make a rent determination based on a fair open market rent, giving them the power to reduce rents if a landlord did not comply with certain existing regulations.

While the Bill technically fell before the May 2021 election, it was taken seriously and instigated the current debate about future rent regulation in Scotland (discussed below). Moreover, the Bill process produced new evidence about attitudes to RPZs and rent regulation in Scotland. This included written and oral evidence to the Scottish Parliament’s Local Government and Communities Committee (LGCC) and overview papers such as Berry and Banks (2021). The tenant union for Scotland, Living Rent, gave evidence to the LGCC highlighting commissioned polling evidence (undated) stating that 75% of respondents called for national rent control for Scotland. This reinforced a 2015 YouGov poll that reported majority support for rent control (LivingRent.org). Recent independent research also found that rent control topped private tenants’ demands across the UK (Soaita, 2022).

Two of the current authors submitted written evidence to the Committee (Gibb & Marsh, 2020). First, they note that a CPI + 1% Scotland-wide ceiling is consistent and transparent, but a blunt instrument that takes no account of market segmentation. Also, setting the ceiling above CPI means landlords would earn real returns and might encourage rent increases to converge on the ceiling. The proposal also raises the possibility of a Scottish allowable rent increase greater than that of the Local Housing Allowance (LHA), that assists low-income renters with their housing costs, which is uprated by the UK Government. The net effects could be perverse: an affordability-inspired measure resulting in deteriorating affordability for lower income households. Second, the proposals for setting fair open market rents (FOMR) raise questions. It is not clear how a FOMR is defined operationally, nor how transparent and accountable Rent Officers would be in making their assessment (this was rather opaque under similar circumstances in the pre-1988 regime). In addition, the proposals interweave price and non-price regulation with the aim of influencing both behaviour and outcomes. We do not understand sufficiently how the proposed changes are likely to interact with the existing regulatory framework. Third, the creation of a database including actual rents for all rental properties operating under the Private Residential Tenancy was, independently, an important proposal that could cut through many of the reported problems with RPZs (Boyle, 2020, for instance notes the misleading reliance on advertised rather than contractual rents). If such comparable, independently-collated rent information were available then the market should function more efficiently. But, to do so, the data need to be consistently collected, easily accessible and fully intelligible.

The then Minister, Kevin Stewart MSP, representing the SNP Government, gave evidence in person to the LGCC (March 3, 2021). While agreeing with the Bill’s aim of eliminating excessive rent increases and promoting affordability, he rejected the proposed approach for three reasons which resonated with the points made by Gibb and Marsh (2020). First, he opposed a national approach that did not account for local variation. Second, he worried that the policy would lead to rents converging on the allowed ceiling. Third, he thought that the rent appeals proposals would swamp the First Tier Tribunal system. He instead proposed reforming RPZs. This position was reinforced with the release of the Housing to 2040 Roadmap on March 15, 2021 (Scottish Government, 2021). The Government stated it would, if re-elected, bring forward a new housing bill early in the new Parliament that would include RPZ reform.

A minority SNP administration was elected in May 2021 and, after protracted negotiations, a form of coalition was created with the Greens. This included making the Greens’ co-leader, Patrick Harvie MSP, minister with responsibility for tenants’ rights with a remit to “implement an effective national system of rent controls” by the end of the current parliament in 2026 (Scottish Government & Scottish Greens, 2021, p26). The italicised statement here is a model of crafted ambiguity which keeps many options open. A clue to where the Green partner in the coalition may be heading can be found in Wightman’s (2020) Scottish Greens policy briefing on RPZs, which concludes (p.13):

‘Scotland needs a method of controlling private sector rents that protects tenants from steep rent rises and incentivises landlords to improve their properties … RPZs would, at best, moderate rents. It would not encourage property owners to improve their properties, [nor] protect tenants signing up to new tenancies. It would also do little to reduce inflated rents.’

At the time of writing, April 2022, the Scottish Government has consulted on a draft rental strategy, including national rent regulation (Scottish Government, 2021). Background work and active evidence-gathering continues in preparation for as-yet-unspecified rent regulation legislation to be brought forward.

Interpreting policymaking on rent control

Reflecting on what is driving recent policy development in Scotland, the Scottish policy trajectory can be read in different ways. The move to RPZs, the case for a national CPI + 1% ceiling, the possible reform of RPZs and the wider discussion now underway, could be interpreted as aligned with the broader trend towards gradually increasing interventionism identified elsewhere (Kettunen & Ruonavaara, 2021). That would align with the broad thrust of the 2016 Act.

However, the mechanisms for establishing RPZs were so unwieldy that it was in practice a symbolic policy: the (then) Government responded to calls to do something about high rents while, in practice, the intervention was ineffective. While this may have been the case prior to the agreement to work with the Greens, the commitment to implement a more impactful policy has been reinforced by the Greens and this is being supported within the wider Housing to 2040 implementation. The Government has created space to work up rent regulation proposals over a period of time, but it currently remains unclear what types of regulatory mechanism will be proposed. Whether the proposals contained in the forthcoming Bill offer a realistic prospect of addressing known implementation weaknesses will assist in discerning the intent behind, and level of stakeholder support for, this agenda or whether it is an example of housing policy “busywork”: “look busy, but change little” (Gurran & Phibbs, 2015).

To what extent is research evidence exerting influence over policy development? Both research and grey evidence has been prominent in the critique of RPZs and also the arguments developed around the private members’ bill. But there is undoubtedly public pressure on the political system to act on affordability, including the activities of new tenants’ rights groups pressing for action on rents, and this has been identified as the key influence on the evolving policy agenda (Slater, 2019). At the same time, perceptions within the policy community that the existing academic literature in economics is, while equivocal, broadly unfavourable towards rent regulation have been scrutinised, including through inviting academic researchers to input directly into policy deliberations (eg. Gibb & Marsh, 2022). At first sight there might appear a potential tension between the policy directions favoured by the campaigns around affordability difficulties, where there is evident support for rent regulation, and the perceived complexion of the conclusions from the evidence base. Yet, as the previous section indicates, it would be possible to legitimate, to use Boswell’s (2009) terminology, whichever position is adopted on rent regulation with reference to the suitable framed account of the evidence base. An alternative reading of this relationship would be that while a decision to proceed with rent regulation might be made on political grounds, that does not preclude the use of the evidence base as a sensitising device to assist in arriving at a policy design that, at least, minimises the risks of negative impacts.

The rent regulation debate intersects with other policy agendas and aspirations. There is enthusiasm in government for creating a positive investment environment for the Build to Rent sector (see Nethercote, 2020). This could create a perceived need for compromise over rent regulation. A further factor is the broader context of ongoing fiscal restraint. For the Government the cheapest way of “doing something” about rental affordability in the short term is rent regulation: all other routes forward require spending public money or condoning tax-expenditures to landlords. Any longer-term costs of rent regulations are precisely that – longer-term – and so might not be weighed so heavily in the decision.

Recent Scottish experience offers a live case study for the role of evidence in housing policy development. Several drivers of policy change are at play: political economy, intra-coalition political bargaining, intersecting policy agendas, and a desire to make policy with reference to the evidence base are all part of the picture. There is considerable scope for further, more detailed exploration of the ways which the relevant evidence base is deployed.

Conclusion

This paper has sought to address three distinct questions regarding the role of evidence in rent regulation policy drawing on both a discussion of the recent academic literature and reflection on recent policy developments in Scotland.

Whatever the substance of the research evidence base, its potential to influence policy needs to be set alongside a diversity of other influences, pressures, and types of evidence. The research on the dynamics of policy change indicates that in practice evidence rarely plays a significant role in policy decisions; other factors have weighed more heavily. The social scientific literature is clear that rent regulation policy, and policy change, can be better understood as a product of the institutional complex of different housing regimes and of political contestation both within and beyond the structures of representative politics.

This leaves open the question of whether, under appropriate conditions, evidence could play a more prominent role in shaping rent regulation policy or does the nature of the evidence base mean that would not be possible? Our review examined the characteristics of the evidence base and the messages that emerge from it. Our key point is that while the evidence can shed light on the impacts of rent regulation, context and contingency are fundamental in shaping policy effects. To apply existing findings directly to guide policy development in other contexts at other times therefore requires making further assumptions about generalisability. We would argue that the literature is better thought of as a device for sensitising to complexity and contingency than for providing simple general conclusions. The way in which the evidence base is brought closer to the policy debate by the grey literature indicates that a range of positions can be sustained. Appeals to the evidence cannot fully adjudicate between these positions. There are irreducible uncertainties: policy making inevitably has to go beyond the evidence base.

Our discussion of contemporary developments in Scotland gives an indication of the drivers shaping the policy process. The policy agenda is being strongly influenced by questions of political economy and political negotiation. Yet, the policy process in Scotland has not simply been cognitively open to evidence but has actively sought to engage with it. The evidence base has been interrogated for the assistance it can give in effective policy design. In this respect, the process displays characteristics of Cairney’s (2019) intermediate positions: we are not in the world of policy-based evidence - decisionmakers value evidence - but the policy process faces imperatives to act that mean the purity of “evidence-based policy” cannot be realised.

We might frame this as a situation of evidence-informed policy, where evidence sits alongside other influences. But it would be more helpful here to differentiate the ‘what’ of policy from the ‘how’ of policy: the evidence base may not have played a substantial role in deciding the broad direction – the ‘what’ - of policy, but once that is determined the ‘how’ can be highly sensitive to lessons from the evidence, including the wisdom of proceeding with caution. This is a refinement of the message from the existing literature, recognising both politics and policy, and acknowledging that the influences on each can differ.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Economic and Social Research Council grant number ES/P008852/1.