Fiscal devolution and the accountability gap: budget scrutiny following tax devolution to Scotland

ABSTRACT Financing regional government involves trade-offs between own-source taxes and grants. Improved accountability has been an argument behind calls for greater tax devolution, but this argument relies upon effective scrutiny mechanisms existing or being developed. This paper explores such issues through the lens of recent tax devolution to Scotland. Drawing on insights from senior stakeholders, we assess how scrutiny has changed in the aftermath of new powers. We conclude that, despite some improvements, progress has been limited. We develop an analytical framework to understand why, drawing out lessons for improving accountability with fiscal decentralization.


INTRODUCTION
Arguments for fiscal devolution have been motivated by a range of factors, including the electoral success of regional parties, pressure from regional nationalism (Brancati, 2008) and territorial discontent with the (lack of) dispersion of national economic gains (Sacchi & Salotti, 2014). The principle of subsidiarity suggests that more local decision-making can, in principle, lead to efficiencies (Oates, 2005), albeit at the expense of economies of scale (Rodriguez-Pose & Gill, 2005).
A perennial challenge in delivering fiscal devolution is that whilst there is a long list of spending areas deemed appropriate for devolution, such as regional health policy and school education, devolution of revenue powers is more difficult (Ter-Minassian, 2015). The result is a 'fiscal gap' or 'vertical imbalance' between what is devolved on spending vis-à-vis revenue, even in highly decentralized countries. Too big a gap can, it is argued, inhibit accountability by not making policymakers face the full fiscal consequences of their decisions (Oates, 1999;Rodden, 2002). Improving accountability through greater tax decentralization has been a key argument used by those in favour of enhanced devolution in the UK and is 'often mooted as perhaps the most significant benefit' (Johnson et al., 2021, p. 20).
Accountability is generally taken to be the 'hallmark of modern democratic governance', a 'hurrah word' (Bovens, 2007, p. 183) frequently invoked but less often demonstrated to have occurred. Others have referred to accountability as a 'complex and chameleon-like' term (Mulgan, 2000, p. 555) and that it has suffered from conceptstretching (Lührman et al., 2020). The literature draws a distinction between vertical and horizontal accountability (O'Donnell et al., 1999) and more recently diagonal accountability (Lührman et al., 2020). Vertical accountability refers to a principal-agent-type relationship between the electorate and the elected (Schedler, 1999). Horizontal accountability refers to how different institutions hold other institutions to account, for example, executives to legislatures. Diagonal accountability is indirect and involves non-state actors, including the role of civil society, academia and the media in holding governments accountable. Our main concern in this paper is with horizontal accountability and in particular relations between Parliament and government, although we also explore connections with issues of vertical and diagonal accountability.
Assessing how fiscal devolution affects accountability requires consideration of the mechanisms that support accountability. A key element of accountability is scrutiny, for it is this that unlocks the information required to hold decision-makers responsible for their actions. How effective this is depends upon the structures, capacities, political contexts and processes that support parliamentaryand widerscrutiny. It is these practical channels that will underpin shifts in accountability, with vehicles of scrutiny varying in their volume and quality. Without these, closing the fiscal gap may not in itself deliver much improvement in accountability, and indeed may worsen it.
This paper contributes to understanding accountability challenges following changes in tax devolution. The experience of Scotland, a devolved nation within the UK, offers a unique opportunity to add to the evidence on fiscal devolution and accountability. The devolved Scottish Parliament has seen a significant increase in tax powers in recent years (taking own-source revenues from around 10% of devolved budgets to nearly 40%). These reforms were, in the words of the cross-party commission established to design them, directly intended to make the Scottish Parliament more accountable 'for the effects of its policy decisions and their resulting benefits or costs' (Smith Commission, 2014, p. 4).
In this paper we focus on the role of scrutiny as a mechanism to support improved accountability following this further fiscal devolution. To do so we draw upon a series of interviews with senior politicians, public officials and external commentators. We supplement these findings with data collected from the public and Members of the Scottish Parliament (MSPs) to explore whether there is evidence that the recent tax devolution reforms have indeed improved scrutiny (in a broader sense) of decision-making and financial accountability.
We find areas of progress, but conclude that scrutiny has become broader and, in many cases, more superficial. Several challenges have emerged whether in the capacity of Parliament to hold the government to account within a more complex framework or in supporting a wider public debate on fiscal accountability. In short, transferring more tax responsibility to devolved institutions to close a 'fiscal gap' on its own may not necessarily improve accountability through more effective scrutiny.
Improving accountability cannot therefore just focus, as it typically does, upon debates over 'how much autonomy' or 'what tax powers to devolve' but also needs to carefully consider reforms to support effective scrutiny. This includes investing in the capacity of devolved legislatures to evaluate and debate more complex arrangements that accompany tax devolution, efforts to improve levels of public discourse on regional fiscal issues, and the establishment of appropriate parliamentary oversight mechanisms. Our research contributes to the literature on fiscal decentralization, fiscal scrutiny and fiscal reform. It also offers timely lessons for countries seeking to embark on greater tax devolution, including the UK (where fiscal devolution remains a source of debate not just in the devolved nations but also in the English regions).
The rest of this paper is structured as follows. Section 2 elaborates on the debate between devolution, accountability and scrutiny. Section 3 summarizes the recent reforms in Scotland and the debates on devolved fiscal accountability. Section 4 sets out our methods, whilst Section 5 summarizes our findings. Section 6 explains the patterns that we identify using four themes identified from the data: politics, processes, capacities and complexities. Section 7 discusses our findings, while Section 8 concludes.

DEVOLUTION, ACCOUNTABILITY AND SCRUTINY
Much of the early literature on fiscal decentralization (or 'fiscal federalism') built from the idea that there were 'efficiency' improvements to be harnessed from local public goods being provided at a more local level, leveraging the knowledge of voter preferences and enabling a closer mapping to public good provision. In short, better quality public services can arise from making regional/local politicians and bureaucrats more responsive to voter demands, in part because public action is closer to citizens (Oates, 1999). Certain public services, such as health, education, regional economic development, and local justice services, all make good candidates for devolving power to subnational governments. But policies susceptible to factor mobility, harmful tax competition (which, of course, depends on the elasticity of the tax base), or large economies of scale (such as many taxes or social security payments), are less likely candidates. Negative tax externalities or concerns over horizonal equity erosion can be important too (Sas, 2017). As a result, a 'fiscal gap' is likely to exist between which spending powers are devolved versus which revenue powers are devolved.
The initial model of UK devolution largely reflected this thinking. In the case of Scotland, own-source tax revenues amounted to just over 6% of devolved spending at the outset of devolutionrising to 10% if local authority tax powers are included in the calculation. A block grant provided the remaining funding. The UK's system of devolution has also always been asymmetric, with different degrees of devolution to Scotland, Wales and Northern Ireland. This has been a challenge to the realization of any 'yardstick' competition benefits anticipated by the literature, with difficulties in understand how differences in outcomes relate to policy choices.
Even in its early stages, however, the fiscal decentralization literature cautioned that devolution of spending powers without accompanying tax powers could lead to inefficient ('Leviathan') outcomes. As this literature evolved, an emphasis on how political and institutional structures could also shape incentives and behaviours amongst voters and policymakers was demonstrated to be key to outcomes (Oates, 2005;Rodriguez-Pose & Gill, 2005). Crucially for our study, this literature identified that a large 'fiscal gap' or 'vertical imbalance' could, in principle, lead to agency problems and inhibit accountability. For example, it argued that in countries with significant subnational borrowing powers, local politicians may end up facing a soft budget constraint by not bearing the full responsibility for the costs of what they choose to spend their money on (Kornai et al., 2003;Rodden, 2004).
Fiscal devolution and the accountability gap: budget scrutiny following tax devolution to Scotland A so-called fiscal illusion may also emerge such that citizens, unable to make a clear link between taxes paid and benefits received, become less likely to sanction regional politicians for poor service delivery (Weingast, 2014). This could lead to a 'flypaper effect' (Hines & Thaler, 1995) where funding from central government grants is used differently (less efficiently) than funding raised through regional taxation. Similarly, it may be difficult for voters to assign accountability to a specific level of government for changes in outcomes which may erode incentives (Geys et al., 2010).
Meanwhile, the literature on accountability distinguished between different types of accountability: vertical and horizontal (O'Donnell et al., 1999) and diagonal (Lührman et al., 2020). In our context, vertical accountability refers to the relationship between the electorate and the elected (Schedler, 1999), while horizontal accountability focuses on the relationship between institutions, and diagonal accountability is a reference to the role civil society, interest groups and the media have in holding governments to account.
Classic accounts of representative democracy emphasize that elections provide accountability (Schumpeter, 1942;Schmitter & Karl, 1991, p. 76). But one of the challenges with vertical accountability is that elections are infrequent, many voters have traditional voting patterns undisturbed by performance, issues may be poorly defined, and clarity of responsibility is blurred with an array of potentially salient issues before the electorate. Elections require information that may not be available and is invariably contested. Effective scrutinyby Parliament and other institutionscan help close this information gap by providing a forum for the executive to be challenged on, and information extracted about, its decisions and resulting outcomes.
Scrutiny is also a key element of horizontal accountability. Specifically, that ministers and officials are accountable to the devolved legislature, in debates in the chamber, parliamentary questions, and examination of executive policies and decisions by parliamentary committees. But this paper also relates to diagonal accountability, in particular through the important role it plays in the 'informational environment' in supporting other forms of accountability. This in turn relates back to, and plays a key role in, vertical accountability, with, as others have noted, voters being more likely to respond when the 'press generates a barrage of new information alerting voters to political corruption' (Chang et al., 2010, p. 215). Crucial to this information environment is the effectiveness of parliamentand wider actors in the policymaking process and environmentto scrutinize fiscal decisions and precipitate the generation and dissemination of information.
It is clear from the literature that accountability must be underpinned by knowledge and information. Oates (2005) noted the emphasis on 'problems of information' in the second-generation theory of fiscal federalism (p. 356). Complexity in institutional design has long been recognized as causing a blurring of responsibility (Hobolt et al., 2013, p. 166). Clarity of responsibility is recognized to be a key factor along with ideological image in understanding the conditions for voting based on performance (Powell & Whitten, 1993). At the same time, who produces what information, when, and why, is crucial. As Rubin and Kelly (2007) note, holding decision-makers to account 'may be difficult or impossible, with the added complexity that the agency being judged is also the one producing the information on outputs and outcomes. The accuracy and transparency of reporting may become critical' (p. 566).
The empirical evidence on the significance of devolution on accountability (or other outcomes for that matter) is mixed (Rodden, 2019). Part of the reason is that crosscountry comparisons suffer from measurement challenges, notably capturing degrees of effective regional autonomy. Identifying a common measure of accountability and/or scrutiny is challenging too, with studies having to rely upon proxies such as measures of policy efficiency (Geys & Moesen, 2009), corruption (Gervasoni, 2010 or governance (Kyriacou et al., 2015). Trends in devolution can evolve too, as highlighted by Bell et al. (2021), often subtly. This makes tracking change difficult without detailed local knowledge. Single country studies offer perhaps the best approach but identifying good case studies to explore these issues is not straightforward (Knight, 2002). For reasons that we will set out, the experience of Scottish devolution provides a good case study to explore issues of tax devolution and budget scrutiny.

SCOTLAND'S FISCAL DEVOLUTION REFORMS
Issues of financial accountability for the Scottish Parliament became an almost constant source of debate in the early years of devolution (Jeffrey, 2007). Whilst increased accountability had been one of the four founding principles of the Scottish Parliament (Scottish Office, 1998), this referred to policy decisions made by the proposed Scottish Executive including on spending priorities within its budget, but not on how this spending was to be financed. As a result, the initial model of devolution created institutions empowered to spend public money but with only very limited, and largely symbolic, powers to raise what it would spend.
Demands for greater tax devolution gained traction on both the pro-independence and pro-UK sides of Scotland's political divide. Both nationalist and unionist politicians spoke of a 'pocket-money' Parliament needing reform (Alexander, 2007;Davidson, 2014;Salmond, 2010). A cross-party consensus emerged that greater tax devolution would 'enhance the democratic and financial accountability of the Scottish Parliament and Government' (Scottish National Party (SNP)-led Scottish Government, 2017), 'create more responsible Scottish politics and would help to remove this grievance culture from it' (Scottish Conservatives, 2014), 'address the issue of a deficit in accountability' (Scottish Liberal Democrats, 2006) and 'enhance the autonomy and accountability of the Scottish Parliament through an extension of tax powers' (Scottish Labour, 2014). The message from politicians, on all sides, was clear: closing the fiscal gap in Scotland would improve decision-making and accountability.
With such a consensus it is not surprising that the level of change in fiscal devolution in Scotland, and the pace at which it has occurred, has been significant. The devolved Scottish Parliament has evolved from a legislature largely focused on how money is spent to one that is responsible for substantial monies raised. These changes have been informed by two reviews: the 2009 Calman Commission and the 2014 Smith Commission. The Calman Commission's remit was: to recommend any changes to the present constitutional arrangements that would enable the Scottish Parliament to serve the people of Scotland better, improve the financial accountability of the Scottish Parliament, and continue to secure the position of Scotland within the United Kingdom.
( UK Government, 2009, p. 1;added emphasis) Ultimately, it recommended a Scottish rate of income taxwith greater flexibility over the power to alter the UK rate of income tax at the marginand devolution of stamp duty land tax (a tax on the purchase of properties) and landfill tax (a tax on depositing commercial waste at landfill). But for many, these reforms did not go far enough. After the 2014 independence referendum, the Smith Commission was set up with an objective to 'strengthen the Scottish devolution settlement and the Scottish Parliament within the UK (including the Parliament's levels of financial accountability)' (Smith Commission, 2014, p. 9; added emphasis). This recommended further powers, including an even greater degree of autonomy over income tax. Since 2016, the Scottish Parliament has been responsible for nonsavings, non-dividend (NSND) income tax (over 90% of income tax revenues) raised in Scotland. The Scottish government has the power to vary income tax rates and bands without constraint, except for the 'personal allowance'. Alongside this, there was further planned devolution of taxes (air passenger duty and aggregates levy) and the assignation of approximately half of Scottish value added tax (VAT) revenues. Figure 1 shows that evolution. Revenues that are assigned rather than devolved are those where there is no devolution of powers to vary these taxes but where an assessment is made of how much of a common UK tax revenue base is raised in Scotland.
Just as in the case of Wales in 2010 and Northern Ireland in 2021, the Scottish debate focused on how the taxes to be devolved and the size of tax take might improve accountability.
There has also been further devolution on the spending side. Eleven currently UK-administered benefits, mainly related to ill-health and the regulated social fund, are in the process of being transferred to the Scottish government.
Once all reforms are implemented, the intention is for devolved and assigned revenues to account for nearly 40% of Scottish government spending. Crucially for the discussion that follows, it is important to note that the process to implement these reforms was negotiated, not between parliaments, but by the UK and Scottish governments leading to the signing of an intergovernmental 'Fiscal Framework' agreement (Eiser, 2020). The resulting regional fiscal model is a mix of a central government-funded block grant, tax devolution, adjustments to the block grant to account for tax devolution, borrowing powers and the creation of new institutions (including an independent budget office, the Scottish Fiscal Commission -SFC).
The result is a 'complex and largely untested' system (Organisation for Economic Co-operation and Development (OECD), 2019). Issues of complexity, and their impact on scrutiny, are a recurring theme in this paper. While arguments continue about whether the desired Fiscal devolution and the accountability gap: budget scrutiny following tax devolution to Scotland level of regional autonomy has been reached, or whether further changes are required, there has been limited focus on what impact this devolution has had on outcomes so far. One reason for this is the lack of a robust empirical investigation of these issues, which is where this paper makes its contribution.

METHODS
To assess whether levels of fiscal scrutiny, and in turn accountability, have improved following tax devolution, and through which channels, we focus upon a series of semi-structured interviews with senior stakeholders involved in the operation of fiscal devolution in Scotland. Alongside this, we undertook a survey of MSPs and another of the general public. This evidence, including crucially access to such influential stakeholders, enables us to establish a deep understanding of the practical experience of operating these new powers from a range of perspectives.
The principal method to gather detailed information on fiscal scrutiny was a series of semi-structured interviews. We interviewed key figures involved in scrutinizing the Scottish budget process (n ¼ 23) in spring 2021. This included senior MSPs (including from Parliament's Finance and Constitution Committee) (n ¼ 7). We also interviewed senior officials from devolved and central government, Parliament and public finance institutions involved in the Scottish budget process (n ¼ 10). We also interviewed external commentators including advisors to Parliament, officials in public bodies (n ¼ 4), and print and broadcast journalists (n ¼ 2).
Research on UK legislatures emphasizes the importance of parliamentary committees (McKay & Johnson, 2010) in the accountability process. Therefore, the Finance and Constitution Committee is crucial to our study. But the effectiveness of accountability extends beyond, including Parliament as a whole and the wider public discourse. Hence, the importance of our interviews with a broad suite of actors.
The interviewees were selected from the researchers' understanding of the public finance and political context. All our interviewees had been in post for several years. Appendix A in the supplemental data online lists the interviewees (with appropriate anonymisation). We ensured that interviewees were from a mix of backgrounds, including all political parties represented in the Scottish Parliament. To ensure reliability, all interviewees were asked similar core questions, which were used as the basis for the data analysis. The discussions evolved depending upon the views of the interviewees enabling a full and comprehensive collection of evidence. Interviews were recorded (where consent was given), transcribed and cross-checked.
We supported this approach with methods designed to add additional insight and colour to our core approach. First, we reviewed key Scottish budget documents, including parliamentary debates, committee evidence sessions and government publications. Second, we alsowith the support of the Scottish Parliament Research Centre -obtained key budget data. Third, we surveyed a subset of MSPs through an online format. Fourth, we commissioned a YouGov survey of over 1000 adults in Scotland.
Our survey of MSPs and the public provided important supplementary context for our research. First, we were keen to better understand the extent to which the public understood the current devolution of fiscal powers postreform and how effective this angle of scrutiny might have evolved. Second, in developing a survey of MSPs we aimed for a broader understanding of their perceptions on how increased devolution has changed scrutiny. Survey responses were received from 27 MSPs (representing 26% of opposition and government backbench MSPs). Whilst this is a self-selected sample, we believe that this is appropriate for our study, particularly given that the strength of this evidence was as much about informing our interviews, and the responses, as it was contextualizing the broader study. A response rate of 26% is relatively high for a study of political 'elites' (Walgrave & Joly, 2018).
The categories that make up our empirical section were drawn from an analysis of both the semi-structured interview data and this supplementary material. Approval was obtained from the University of Strathclyde's Economics Department Ethics Committee.

BETTER SCRUTINY AND IMPROVED ACCOUNTABILITY?
In this section we explore whether the extent, form and effectiveness of scrutiny of budget issues undertaken in Scotland has improved following the transfer of tax powers. Our conclusion is best articulated by a senior member of the Scottish Parliament's Finance Committee, who told us that 'it has become broader, but also more superficial' (interviewee 1). Throughout our interviews with senior MSPs and policymakers the consensus view was one of improvement, but ultimately disappointment in how scrutiny had evolved. Whilst policymakers are more aware of both sides of the balance sheet post-tax devolution, improvements in scrutiny have been patchy (interviewees 7 and 18). One senior MSP articulated: 'progress has been made but it is hugely frustrating; the hope had been that [the] Smith [Commission's reforms] would have moved us much further forward' (interviewee 1). Some had stronger views with one suggesting that the ability to hold government to account had 'got worse' (interviewee 1) whilst another MSP lamented that the capacity for scrutiny had been 'squeezed' since tax devolution (interview 2).
These views of senior MSPs immersed in budget scrutiny are broadly consistent with the views of parliamentarians as a whole. In our survey of MSPs, 55% responded that scrutiny had only 'got a bit better' with 30% saying it had 'stayed the same' or 'got worse' since the new reforms. None said it had 'got a lot better'. The nature of scrutiny is multifaceted, but on basic metrics of parliamentary interest in budgetary matters, there is only partial evidence of a sustained change in scrutiny. One public servant reflected that 'surprisingly, the number of parliamentary questions on the Budget has not really increased. We were expecting a lot more' (interviewee 13). Another highlighted that the time spent in committee scrutinizing 'strategic issues had fallen, in part because of a squeeze on time but also other priorities' (interviewee 9). Data provided by the Scottish Parliament for the purposes of this research do show an increase in the number of budget-related questions (written and oral) laid in Parliament, but this has fallen back recently (Table 1).
Clearly, volume is only a partial measure of scrutiny but offers an insight into the day-to-day rhythm of parliamentary time. It is possible, of course, that other factors have helped to squeeze budget debates, including Brexit. But our interviewees were, on the whole, of the view that whilst there might be improvements over time, 'blockages' were more structural with one MSP 'pessimistic about potential improvements over time' (interviewee 4). Concerns were raised about the depth of scrutiny undertaken in parliamentary debates and committee evidence sessions. We were told, for example, that day-to-day efforts around evaluating the budget were 'shallower and more cursory than when we had a simpler budget [based upon a block grant]' (interviewee 2). A particular outcome since tax devolution reforms was a concern over clarity of focus. Considerable energy appears to have been spent scrutinizing the technicalities of the new rules that govern fiscal devolution, as opposed to the policy choices or outcomes obtained (interviewee 18). This view was reinforced by a senior public servant: 'Parliament is pressing more on budget issues. But the question is whether people are focussing on the right things. Lots of focus on intricacies of the new framework, but no one asking the big questionse.g., economic performance?' (interviewee 10). We found evidence that those involved in the budget process felt that their efforts to scrutinize fiscal issues remained 'worthy but not headline-grabbing' (interviewee 6).
One area of improvement has been scrutiny of the new powers devolved to the Scottish Parliament. Such debates have been of 'good quality' (interviewee 3). But interviewees spoke of parliamentarians taking their 'eye off the ball' (interviewee 6) on broader areas of the Budget, and in particular scrutiny of spending. Debates have tended to be superficial, limited to whether a portfolio spending line is going up or down (and by how much), or whether existing tax rates are higher or lower than in the UK (interviewee 19). A senior public servant concluded that 'whilst there is more interest in the budget at a strategic level, there's little evidence that there has been a stepchange in strategic thinking around financial accountability' (interviewee 8).
We explored whether changes in parliamentary scrutiny had facilitated broader efforts at improving scrutiny via external organizations and in turn the public. In the cross-party Commission's report that recommended more fiscal powers for Scotland, the Chair Lord Smith of Kelvin wrote that 'a challenge facing both Parliaments is the relatively weak understanding of the current devolution settlement' (Smith Commission, 2014, p. 6). External groups largely reinforced the insights from the parliamentarians we interviewed. As one journalist stated: 'we have a responsibility to cover the Scottish Budget, but beyond headline tax announcements it is so hideously complex to articulate what is going on that it's nigh on impossible to engage the general public' (interviewee 23). There are some signs of greater engagement with new investments in academia and think-tanks leading to increased commentary on budget issues but this remains in its early stages.
In summary, whilst there has been an improvement, it is hard to conclude that the devolution of tax powers and the resulting closing of the devolved fiscal gap has brought the transformation in accountability that had been hoped for. Whilst there have been some positive developments, these have been limited to specific areas. On reflection, what is striking in the Scottish case is the presumption that the simple act of devolving tax powers in itself would lead to more effective scrutiny. The reality has been different.

A FRAMEWORK TO UNDERSTAND THESE DEVELOPMENTS
Having set out the experience of these devolved powers, and specifically the conclusion that there has only been a limited step-change in fiscal scrutiny post-tax devolution, we develop an analytical framework to explain such findings. This framework has four pillarspolitics, complexities, processes and capacitieswhich, we argue, have shaped how the process of fiscal scrutiny and accountability has evolved.

Politics
The first pillar to shape the evolution of regional fiscal scrutiny is the political context. Bell et al. (2021) highlight how the design of any fiscal framework for regional governments tends to evolve out of an ad hoc political process as opposed to a principles-based approach. This invites us to consider the underlying environment in which powers have been devolved. Are they part of a nationwide review of federal structures or are they a political response to demands for greater regional autonomy (including possible secession) driven by regional party electoral success? In essence, are debates on 'more powers' framed in governance and public policy terms or autonomist/nationalist vs Fiscal devolution and the accountability gap: budget scrutiny following tax devolution to Scotland centralized power terms, even acknowledging that these are not necessarily mutually exclusive? In the case of Scotland, the answer is that debate is framed in terms of wider constitutional rather than governance terms. As outlined above, issues of improved accountability had been used as an argument for more powers, but the reality is that the wider constitutional context has dominated most aspects of policymaking in Scotland in recent years.
The political context has meant that the devolved fiscal framework itself has often become a source of tension between the Scottish and UK governments. As a result, rather than providing greater clarity over responsibilities to help promote fiscal scrutiny and better accountability, budget debates have been framed within an environment of considerable 'political noise' (interviewee 19). The Scottish government, for example, has repeatedly emphasized in debates over its budget choices (including when defending any unpopular cuts to programmes) that UK government 'austerity' is being imposed on its block grant (Forbes, 2020) and of the effects of Brexit on its devolved tax revenuesthe SNP-led Scottish government opposed the UK decision to leave the European Union (EU) (Scottish Government, 2020). The UK government has criticized the management of the Scottish economy by the Scottish government and the impact on devolved taxes (Daily Mail, 2019). In effect, the transfer of tax powers has been used on numerous occasions as a 'political football used to score points, a feature which is as predictable as it is depressing' (interviewee 20).
This reflects what Chris Hood identified as a key challenge with multilevel governance, namely that 'systems of divided government and complex multi-level government structures may offer more opportunities for blame-shifting in the sideways and up-and-down directions than simpler structures' (Hood, 2011, p. 43). This 'blame-shifting' has been aided by the complex set of arrangements to enact fiscal devolution in Scotland. It perhaps explains the limited appetite to simplify these arrangements, for example, to agree on a set of common budget facts each year. The result, however, is that the political noise has often tended to dominateand crowd outefforts to improve fiscal scrutiny.
There are also wider devolved party-political dynamics that have played a role. UK devolution was seen by some as an attempt to move to a new form of politics, away from a more majoritarian form of democracy to a model of consensus and collaboration. The early hope of 'new politics' included notions of creating more accountable and crossparty politics (Paterson et al., 2001) but critics have argued that, in reality, Scotland has ended up with the attributes of a Westminster-style Parliament (Mitchell, 2000;Cairney & Widfeldt, 2015). Party discipline on key policy issues has been a key feature of politics in Scotland (Mitchell et al., 2011). In this respect, budget scrutiny at Holyrood has ended up little different to how it takes place at Westminster. This culture has arguably acted as a constraining device upon fiscal scrutiny, with many MSPs tending to 'toe the party line' (interviewee 2).
Moreover, several MSPs identified that there was a reluctance to 'talk about controversial stuff' (interviewee 3). This unwillingness to explore potentially contentious financial issues across party-political boundaries has shaped how committees have approached budget scrutiny. A senior public official told us 'a key objective has been to turn the focus away from political bunfights and more toward technical aspects of accountability' (interviewee 9).
The political context is therefore crucial to how reforms shape the focus and tone of the debate around scrutiny. In Scotland, the focus on wider constitutional debates coupled with strict party discipline appears to have acted as a constraint on the degree of rigour in pursuing issues of fiscal accountability.

Complexity
Enacting any sort of fiscally decentralized structure will involve complexity, with that degree of complexity typically a function of the extent of tax devolution rather than that of spending. Boex and Kelly (2013) set out a series of principles to govern an effective fiscally federal arrangement: '(i) a clear assignment of expenditure responsibilities; (ii) some own revenue sources, (iii) an effective intergovernmental fiscal transfer system; and (iv) a framework for local borrowing ' (pp. 260-261). Each of these components is present in Scottish devolution, and each has changed with the recent reforms but, often, in an ad-hoc way. The effect of which is an increasingly complex system, with limited assessment or awareness of the effectiveness of the way each of these principles operates in practice. The complexity with which further devolution has been implemented comes through strongly in our analysis as the most powerful perceived barrier to improving scrutiny.
The current devolution settlement means that the Scottish Budget is a mix of a block grant, devolved and shared tax revenues, block grant adjustments, borrowing programmes and rules on the use of savings (the Scotland Reserve). Complex rules now govern how regional tax forecasts are 'reconciled' with outturn data, with the effect that budget changes in one year feed through to the amount the government has to spend many years into the future. On top of this, no one institution is responsible for communicating all the facts on these different components, with HM Treasury, HM Revenue & Customs (HMRC), the Department for Work and Pensions (DWP), the Scottish government, Scottish Fiscal Commission (SFC), and the Office for Budget Responsibility (OBR) all having a role. This system is woven on top of an already complex and asymmetric system for the funding of the regions and nations of the UK.
This inherent complexity was not discussed in depth during negotiations over the transfer of powers. Indeed, one official involved in the negotiations told us: 'The principal challenge, which I'm not sure anyone fully appreciated back in 2014 [when the reforms were being designed], is the complexity in the framework itself. Difficult to explain and difficult to understand' (interviewee 15). These reflections were shared amongst politicians too: 'I am a huge supporter of more powers for Scotland, but I'm pessimistic about the future of Scotland's Budget framework, it's just too horribly complex' (interviewee 4).
This complexity has meant that significant resources and time have had to be invested in working through the complexities of the framework each year. A public servant leading on budget issues remarked: In reality, a lot of our time gets devoted to understanding the small-scale specifics of the framework, and what is happening to them, so the big picture gets lost. The UK system is not perfect, but there's a much sharper relationship between spending, revenues and budgets that allows for better scrutiny of what decisions mean.
(interviewee 10) The origins of this provide important lessons. The consensus-building approach that was integral to the Smith Commission's work meant that each political party's 'red lines' were embedded in the final deal. Reforms were negotiated, agreed and designed in fewer than two months. As a result, detailed consideration of the overall implications of the new arrangements and crucially how the new powers would be operationalized were squeezed. The result was, unsurprisingly, 'a hugely complex framework on top of an already asymmetric and unique devolution system in the UK, within a public finance framework at the UK level which is already hugely complex and opaque' (interviewee 4). Many of our interviewees spoke of the need for simplification, with failure to do so potentially creating issues of trust in devolution itself (interviewee 1). Others suggested that the complexity suited the political objectives of both governments (interviewee 2). In our survey of MSPs, on a scale of 1 (very low) to 10 (very high), we asked them to assess the level of understanding of fiscal issues in Parliament. The average score was 3.5. To what extent is Scotland unique here? It would be naïve to expect significant numbers of politicians to take a keen interest in fiscal scrutiny all the time. But the situation in Scotland appears 'particularly challenging' (interviewee 18).
A similar picture emerges when looking at levels of public understandingproviding some insight into vertical accountability. One journalist remarked that: one of the biggest challenges with this framework is that simple questions that most people would want to know the answer tofor example, has a different income tax policy had a positive or negative impact upon the Scottish Budgetimmediately get caught up in a technical debate about adjustments, reconciliations and forecasts. It's impossible to make it relatable.
(interviewee 22) The complexity of the reforms is shown up in levels of public understanding of Scotland's devolution arrangements pre-and post-reform. In our survey of 1000 adults, we found a higher level of understanding of the responsibilities within the original 1999 devolution settlement as opposed to the recent 2016 tax devolution reforms. For example, over 70% of those surveyed correctly identified the Scottish government as being 'most responsible' for the National Health Service (NHS) in Scotland a key aspect of devolution in 1999. An even higher figure was found for schools policies (83%), which was also an early aspect of devolution. Likewise, 68% correctly identified the UK government as 'most responsible' for foreign affairs, a policy that has never changed in the period since devolution. In contrast, on the 2016 reforms, nearly a fifth, however, did not know who is most responsible for income tax (37% amongst 16-24-year-olds), with a quarter still believing (wrongly) that the UK government is 'only' responsible, despite this not having being the case for five years. Only half of voters reported even being aware that the Scottish Parliament's powers over taxation had increased in recent years, and this is despite the biggest change to devolution in two decades.

Processes
Studies have shown that as responsibilities shift, it is important to design effective processes to support fiscal scrutiny (Wehner, 2014). This can range from investing in new data sources to aid intelligence gathering through to new institutions to hold policymakers to account.
Mindful of this, new arrangements for analysing Scotland's economy and public finances have indeed been established, including an independent forecaster: the SFC. There have also been changes to the day-to-day oversight of the budget by Parliament (Gardner, 2019;Scottish Parliament, 2017). In particular, efforts have been made to extend year-round scrutiny of budget decisions. This new processin principle -'pushes all the buttons' (interviewee 2). The creation of new independent institutions, such as the SFC, is also seen as 'a good thing' (interviewee 3).
But processes have lagged in other areas. Despite the significant devolution of fiscal responsibilities since 2016, the lack of a dedicated parliamentary committee focused solely on budget issues is a significant oversight (interviewee 1). From 2016 to 2021, fiscal scrutiny was part of a committee examining constitutional issues. This is believed to have had two effects. First, it made the politics of fiscal scrutiny more factional (for the reasons set out above). Second, it constrained the amount of time MSPs could devote to discussing fiscal issues.
Moreover, there appears to be a general lack of agreement about the allocation of responsibility for scrutinizing budget outcomes across committees in Parliament or the process for doing so. The majority of Scottish Parliament committees are segmented by portfolio (e.g., health, justice, etc.). These focus on individual areas and do not, typically, look at cross-cutting issues or overall fiscal sustainability. But equally, Finance and Constitution Committee members told us that issues of spending are a responsibility for portfolio committees, so they were reluctant to scrutinize spending decisions in detail for fear of cutting across the work of colleagues. The result can be a limited focus on strategic spending questions and, as noted by Parliament's Budget Process Review Group, a tendency to scrutinize budgets in silos with cross-cutting Fiscal devolution and the accountability gap: budget scrutiny following tax devolution to Scotland issues falling through the cracks (Scottish Parliament, 2017). In addition, how committees scrutinize their portfolio budget allocations (and how much time they take) varies depending upon the interests of members and workload. A Scottish Parliament committee convenor explained to us: 'My Committee was responsible for scrutinizing two highly complex bills in this Parliament. The amount of effort that was required for that was enormous. Budget scrutiny barely registered in our thoughts' (interviewee 2). As a result, 'big strategic issues fall through the cracks' (interviewee 3).
Aside from parliamentary processes, there is a sense that levels of budget information have improved, but that more could be done to keep pace with efforts to support a step-change in scrutiny. Budget documentation was singled out as having been 'a bit of a problem' (interviewee 2) and that it was 'impossible to find what you are looking for' (interviewee 14). A lack of 'whole of government' accounts or a record of actual spending (as opposed to planned spend) are important gaps. As a senior public official told us, 'on economy and taxation, most information is available, but it can be hard to pull together in a way that makes sense to people' (interviewee 13).
A related issue reflects the timing of the budget process. Under the new Scottish tax devolution system, the amount of funding that the Scottish government has each year is a mix of both own decisions on devolved taxes, but also decisions by the UK government on the block grant to the Scottish government and, importantly, decisions on English taxes equivalent to the Scottish devolved taxes (which determine how Scotland's block grant should be adjusted year on year). But, in nearly every year since 2016, the timing of these UK decisions has been uncertain and late. It cannot be overestimated the challenges that this timing issue has had on fiscal scrutiny. Often, draft budgets have had to be prepared conditional on further details being provided at a later date. This came to a head in 2021 when differences in the timing of forecasts made by the two independent forecasting institutionsthe OBR and SFCled to additional flexibilities in funding being made available simply because of when these two institutions published their reports.
In part, this reflects the asymmetric nature of devolution in the UK. Unlike federal countries where budget arrangements between central and subcentral governments tend to follow a standard process, in the UK the approach has been reactive, iterative and largely ad hoc. As one external commentator pointed out: If the UK Government was serious about making devolution work then it would make sure to choreograph budget timings and reflect upon how its decisions on budget timetables impacted upon the devolved administrations. This is what happens in every other country that has a significant degree of regional autonomy.
(interviewee 20) Lastly, a challenge highlighted by some appears to be the lack of oversightor refereeof Scotland's budget framework. In total, seven different institutions play a role in determining how much funding is available in the Scottish budget each year. In our survey, 64% of MSPs agreed with the statement: 'Some suggest that there is a need for a single independent body/entity to oversee the Fiscal Frameworkakin to a "referee". Do you agree or disagree with this statement?'

Capacity
A final aspect of our framework to understand what drives budget scrutiny is the underlying analytical capacity of a devolved region to take on new powers. There are several themes to this: timing, knowledge and outside skills. In our case study, only a minority of MSPs responding to our survey believed that there was sufficient time and resources to enable them to scrutinize the budget properly. Around a third believed that there was both 'insufficient time and insufficient resource'. Only 13% believed that there was 'sufficient time and resource'.
This was a theme supported by data collected from our interviews: The biggest constraint is a lack of time. Efforts to create a year-round Budget scrutiny process haven't really worked so far, everything still gets squeezed into a very short timetable, made all the more difficult by recent decisions of the UK Government to delay their Budget.
(interviewee 4) These concerns can be particularly pronounced when there is a high turnover of members on key budget scrutiny bodies. Between 2016 and 2021, the Finance and Constitution Committee, with a standing membership of 11 at any one time, had 29 different members. In addition, the level of resources invested in capacitybuilding to support fiscal devolution is important. During discussions over the transfer of new powers to Scotland, there was limited debate about what skills would be needed, outside of technical roles required to implement the new tax and social security responsibilities (interviewee 20). As a result, there is a 'real asymmetry of power, not just over information, but resources' (interviewee 2). The level of investment in parliamentary and wider public resources has arguably not kept pace with the degree of autonomy. Seven full-time equivalent (FTE) staff are employed by the Scottish Parliament's Scrutiny Unit to support MSPs scrutinize an annual budget of over £40 billion.
It is not just within Parliament where it is important to develop capacity. The level of understanding of Scotland's budget framework appearsso farlimited beyond a small number of institutions and academics: There have been welcome improvements in the fiscal community from the Scottish Fiscal Commission through to a network of academics across the UK working in devolved public finance issues. But I still have a nervousness that there are not many people who know this stuff.
(interviewee 12) There is a concern that knowledge of the fiscal framework, for example, is common to a few individuals. Should they move on, knowledge will be lost. More generally, there is a sense that Scotland lacks enough academics and think tanks engaged in such debates to scrutinize and interpret budget issues. There is a similar lack of strength in the number of journalists working in this area with posts cut in recent years.

POLICY IMPLICATIONS
A key finding from our research is that the significant changes in the Scottish Parliament's fiscal competencies have not been complementedthus farby equivalent improvements in financial scrutiny. We identify only limited progress in improving the scrutiny of fiscal issues, despite this being a key motivation for tax devolution in the first place. There have been more informed discussions of tax policy choices to pay for public services, as one might expect. But this appears to have come at the expense of 'depth' and has been accompanied by the creation of a complex framework of which there is only limited understanding. The consequence is that fiscal scrutiny has arguably become less detailed and incisive, particularly on important structural questions. We set out four main explanations for this in the previous section: politics, complexities, processes and capacities. While each of these is, in themselves important, they are also linked and raise important wider lessons for the UK and countries considering greater tax devolution. Where negotiations over the transfer of new powers are largely a government-to-government initiative undertaken in a highly politicized environment, the focus of attention is likely to be the transfer and operation of the new responsibilities rather than the changes required to provide adequate scrutinyin both Parliament and beyondto make the ambitions of greater fiscal accountability a success. It is vital therefore that reforms of parliamentary institutions are placed at the heart of debates on the devolution of fiscal powers. It is important that such discussions take place before the transfer of powers are negotiated. This avoids reforms being bolted onto existing structures as opposed to being carefully designed from the start.
Consideration should also be given to the level (and form) of resources needed for effective fiscal scrutiny (both inside and outside Parliament) so that they match any growth in tax powers. The Scottish experience has shown the importance of agreeing in advance approaches to practical issues, such as the timings of when budget decisions are made in a system where there are interdependencies and spillovers, to provide clarity to the policy process. Any increase in power puts added time pressures on parliamentarians, public bodies and journalists to process the raft of information required, with a failure to plan for this only likely to inhibit public understanding and debate. In the Scottish case, such issues have only been partially addressed, with officials, MSPs and journalists needing to pore over hundreds of pages of documents in a short time frame. The Scottish-specific budget system has been appended onto a Westminster system of budgetary policymaking that was already viewed as being toward the least effective end of the spectrum of liberal democracies (Wehner, 2006(Wehner, , 2014. Efforts to improve the arrangements for fiscal scrutiny after the devolution of further powers cannot be expected to attract the same degree of commitment as those agreed to as part of the devolution of these powers. This came through clearly in the case of fiscal devolution to Scotland where efforts to improve year-round scrutiny undertaken by Parliament have had mixed success, whilst in contrast, aspects of the framework formally agreed upon as part of the government-to-government agreement of the fiscal framework, such as the establishment of an independent SFC, have been more successful. A key lesson, therefore, is the need to focus not just on the institutional structures and levels of investment required to implement new powers (and to hope that accountability improvements follow) but to think carefully about the broader structures and investments requiredin Parliament and beyondto deliver greater accountability.
Perhaps the biggest barrier that fiscal devolution in a modern economy faces is the inherently highly complex nature of the fiscal system itself. This is particularly the case where the approach lacks the kind of principledbased approach advocated by Bell et al. (2021). It is this that lies at the heart of many of the practical challenges for scrutiny and in turn different forms of accountability. Again, this offers important lessons for reform. Complex systems designed to be 'all things to everyone' can, particularly in the context of an already complex system of fiscal devolution, make the very objective that is being sought by these reforms more difficult. Democratic institutional design requires consideration of scrutiny as a central consideration from the outset given its key role in supporting accountability. It cannot be left as an afterthought.

CONCLUSIONS
A rich literature exists on whether fiscal decentralization can lead to better outcomes. Such dividends can come through several channels but transferring powers to regional governments can also entail costs and new demands on institutions. One strand of this literature has been on whether tax devolution and the narrowing of any 'fiscal gap' between spending and revenue responsibilities, can improve accountability. The current programme of fiscal devolution to Scotland was motivated by some of the core conclusions of this literature. But a key feature of the recent literature is the importance of supporting institutional and governance arrangements that accompany the transfer of more powers.
In this paper we have explored the development of fiscal scrutiny in the context of recent tax devolution to the Scottish Parliament. Given that the transfer of these fiscal responsibilities has only been partially completed, our analysis should be viewed as an assessment of 'work in progress'. Revisiting our conclusions at some future date will be important to assess the longevity of the effects that we Fiscal devolution and the accountability gap: budget scrutiny following tax devolution to Scotland have identified. At the same time, future research might helpfully explore in greater detail the different complexities of accountability within a multilevel system such as the UK. Our research has provided a case study of fiscal devolution within the UK, but there are other dimensions worth exploring including issues of local government accountability and unpicking the interactions and overlaps between different formsvertical, horizontal and diagonalof accountability within the context of regional fiscal policy.
In exploring how scrutiny has evolved thus far, our evidence suggests that some of the anticipated accountability benefits of tax devolution are being realized, but that hopes for a significant step-change in fiscal scrutiny and accountability have only been partially realised (and in some cases disappointed). Improved accountability had been presumed to follow increased tax devolution, and politicians across all the main parties in the Scottish Parliament used the rhetoric of accountability as a 'hurrah word' (Bovens, 2007, p. 183) uncritically. But what has become clear is that attention needs to be paid to developing effective scrutiny mechanisms to support vertical, horizontal and diagonal accountability.
Several technical challenges, from the timing of UK government budget announcements through to the quality of the information provided, have made scrutiny more challenging than before tax devolution. The political context following the independence referendum in 2014 has also been a factor, with parties increasingly polarized over Scotland's constitutional status. But more strategically, when plans for greater devolution were being drawn up there was arguably a failure to consider what devolution of new tax powers would mean for the capacity of parliamentarians (and others) to scrutinize a more complex budget process with greater tax devolution.
On reflection, the focus of debate in Scotland was largely upon the scale of tax devolution as a means of improving accountability, with less recognition of the underlying institutional and governance framework in which these tax powers would subsequently sit. Attempts to address this, through different review groups, have made some progress but will take time. Scotland's experience offers some important insights for other regions seeking to take on more devolved tax powers. Greater tax devolution can bring with it improvements in accountability. To be fully realised, however, careful consideration of not just 'what' to devolve is needed, but of the processes that underpin that transfer of power.