INSTITUTIONAL GEOGRAPHY ONCE MORE? DEVOLUTION ECONOMICS NORTHERN IRELAND STYLE

Abstract

In this paper, we investigate devolution in the Northern Irish context. The role of institutional geography is highlighted. Both contemporary and historical (c. 1921–1972) devolution models are discussed. It is argued that institutional factors provide explanations for economic weaknesses found under both devolution models. The evidence presented suggests that the institutional geography has both changed and exhibited continuities and we discuss these features in relation to contemporary issues. In particular, it is argued that institutional geography considerations have shaped the creation of the current review of tax devolution as well as the creation of an independent fiscal council.

1. Introduction

Recently, there have been attempts at connecting analysis of devolution with the long-standing productivity weaknesses and regional economic unevenness in the UK. The conclusions of these papers tend to make connections between political institutions and unsatisfactory socio-economic outcomes (McCann, 2021; Pabst and Westwood, 2021). In this paper, we investigate themes including governance, political economy and institutions in considering the economic consequences of the Northern Ireland (henceforth NI) experience of devolution over the long run as well as offering some contemporary insights regarding public finance and fiscal devolution in NI.

The possible interconnections between extending political decentralisation and reversing long-standing regional economic and social inequalities have come into UK-wide public policy debate with the recent (February 2022) White Paper on Levelling-up (HM Government, 2022). In the 432 page document, 12 missions are identified to act like anchors for policy, analogous to the way net zero is supposed to anchor efforts to decarbonise the UK economy (Swinney, 2022). It is for instance promised that every part of England that wants a devolution deal will have one and that local government funding pots will be simplified. While the White Paper promises ‘root and branch reform of government and governance in the UK’, this will be easier said than done. Furthermore, in order to ensure levelling up works for the whole UK, it is necessary to think about how for instance raising the level of educational attainment can keep pace with spreading the supply of high-skilled, high-paid jobs. While the economic divides discussed in the report stretch back at least a century, the stated achievement of its missions by 2030, implying reversing regional inequalities, appears an unrealistic target, particularly with funding only outlined until 2025 (Swinney, 2022).

The concept of institutional geography is to the fore in this paper, and it offers a deeper analysis of the economics of devolution. At root, the concept can be defined as indicating that the costs and benefits of a particular devolved settlement rest on the particular institutional mechanisms under which devolution occurs; furthermore, these institutional settlements in turn are shaped by historical considerations (Rodríguez-Pose and Gill, 2005, p. 407). As Rodríguez-Pose and Gill (2005, p. 407) state in their seminal paper:

Given that different forms of devolution hold very different economic implications, the relative influence governmental tiers have over the form and content of the decentralization process becomes an important determinant of the economic benefit, or cost, that devolution is likely to bring about.

The institutional geography perspective is a ‘second generation’ type response to the overly optimistic predictions of the ‘first generation’ fiscal federalism literature. In the ‘first generation’ literature, decentralised governmental systems offered economic dividends (Hamlin, 1991; Oates, 1972, 2005; Tiebout, 1956). ¹ Furthermore, in the ‘first generation’ literature movements of people (as in Tiebout) and/or fiscal resources would bring about more optimal spatial allocations.

Institutional geography provides a more realistic way of looking at spatial economic activity, and it recognises that spatial economic disparities can persist and that political decentralisation might close these gaps. Germany has after all spent €2 trillion on reunification costs with mixed success and this despite substantial internal migration away from the former DDR (Swinney, 2022). Despite political decentralisation, flows of money and inter-regional migration that should have promoted convergence, according to ‘first generation’ type arguments, a gap remains. Institutional geography suggests that even within jurisdictions, such as Germany or the UK, institutional structures may continue to differ in ways that at the very least reinforce spatial inequalities. The economic dividends that are supposed to automatically flow from political decentralisation need not, because the world does not look like a ‘first generation’ model. This line of argument is inherently contextual and location specific in its implications: it implies that different decentralised or devolved settlements may have different economic consequences in the one location or that similar sets of policies can have different outcomes depending on the institutional configuration.

The specific experience of NI reinforces the relevance of a more general institutional geography approach to understanding the political economy of UK devolution. It was in the Government of Ireland (GOI) Act 1920, in language borrowed from Acts granting autonomy to New Zealand, Canada and Australia, that for the first time within the UK that autonomy was to be granted to another parliament (Torrance, 2020, p. 17). Under the GOI 1920, a reserved powers model was followed. In following such a model, the Act listed which matters were to remain the UK government’s (or ‘Westminster’s) responsibility (excepted or reserved) with everything else devolved (or transferred) to the Parliament and Government of NI (Torrance, 2020, p. 17). ²

While some specific financial features of the GOI 1920 did change over time, for instance the Imperial Contribution with the associated Joint Exchequer Board was abolished in 1973 and replaced with a move towards a Block Grant model. This financial reform was in turn was modified further by the creation of the Barnett formula. ³ However, the underlying philosophy in NI remains a reserved powers one (Torrance, 2020, p. 148). While Scotland and Wales moved quicker to embrace fiscal devolution, even in Scotland the pace of reform has often been somewhat glacial. As late as 2014, prior to greater tax devolution in Scotland, the underlying reserved powers philosophy was memorably described by Armstrong and Ebell as offering a ‘pocket money’ form of financial settlement in which devolved governments were mostly tasked with spending an allocated funding amount. Even the terms of the Scotland Act 2012 limited the scope for moving beyond a balanced budget (Armstrong and Ebell, 2014, p. F5).

The remainder of the paper is structured as follows. In the following section, we observe the more general insights that might follow from examining devolution in NI. We then outline the economic predicament of the region and note that three different sets of processes are at work. From that we move on to discuss the historical institutional geography of NI in more detail. We then reconsider Heald’s (2003) important study of public finance and devolution and his analysis of the then prevailing institutional geography. We observe the changes and continuities that have been experienced since Heald’s report. In the final section, we consider how considerations of institutional geography are implicit in the work of the recently created Northern Ireland Fiscal Council and the Independent Fiscal Commission.

2. Northern Ireland: A place apart?

It has been demonstrated that the introduction of the Barnett formula was a consequence of Westminster’s previous financial relationship with NI (King and Eiser, 2016; Torrance, 2020, p. 167). Given that the legislative structures in NI, and the associated approach to financial transfers pioneered in the 1920s, continue to influence not only its own financial machinery a century later but also the Welsh and Scottish settlements established in the 1990s, and modified since, it is unfortunate that connections between NI’s economic and public finance performance do not account for a greater emphasis in the wider literature considering the political economy of UK devolution. Recent work however is better in this regard as it considers how productivity has related to UK devolution (McCann, 2021; Pabst and Westwood, 2021).

Even as eminent a scholar of the economics of political decentralisation as Barry Weingast in his 1995 discussion of UK experience, while he did discuss the English experience during the eighteenth century, as part of his discussion of what he termed ‘market-preserving federalism’, fails to makes any reference to the far more overtly federal basis of the GOI 1920 much less the extent to which it did conform to a market-preserving model (Weingast, 1995). We hence get no evaluation of the contribution of devolution in thinking about the economy of NI much less do we get a discussion of the extent to which market-preserving elements are relevant in discussing the way devolution operated in the five decades before the Stormont Parliament was prorogued. There are general insights that can be learned from analysing the devolved settlements that existed c. 1921–1972 and since the 1990s. The economic history of NI should not just be equated with the political turmoil that ensued during the Troubles. ⁴ While many of its economic problems are long standing, and while it is often assumed that these problems are the result of its uniquely violent past, it is the case that the region’s economic predicament predated the Troubles.

Political settlement is an important component of the contemporary institutional geography of NI. It would be misleading to ignore the economic benefits of political stability (Teague, 2021). More than a quarter of a century after the 1998 Good Friday/Belfast Agreement, the regional economy looks considerably better than it did during the Troubles (Jordan and Turner, 2021). NI by the 2020s has some clear economic strengths, such as its ability to attract inward investment, its ICT, cybersecurity and Fintech clusters, the thriving film and TV industry and the related strengths of its best educated students, to name a few (Brownlow, 2020; Independent Fiscal Commission, 2021, p. 33). Life satisfaction and happiness data also reflect that NI does well relative to other UK devolved nations and English regions (Independent Fiscal Commission, 2021, p. 33).

However, while peace has led to undoubted economic benefits, even the attainment of political stability as a result of the 1998 Agreement has not proven to be the end of economic weaknesses that often predated the outbreak of violence in the late 1960s. As has been noted by a range of economists, significant economic challenges have persisted despite the 1998 Agreement (FitzGerald and Morgenroth, 2020; Teague, 2021). On whatever metric chosen, NI has consistently been one of the UK’s most poorly performing regions and its economic turnaround has not narrowed the gap with better performing regional economies in the UK. ⁵

Furthermore, it has been suggested that the growth of public expenditure, which emerged in response to the conditions underlying civil unrest, as well as providing part of an interrelated strategy response to violence, has not been fully reversed by political agreement. Indeed, maintaining political stability has led to the persistence of a large regional public sector despite calls for rebalancing towards a larger private sector (Brownlow and Birnie, 2018). ⁶ Overall then, the economic promise of a ‘peace dividend’—interpreted as representing a shrinking of the public sector role in the economy—discussed back in the 1990s, remains an unfulfilled one more than two decades after the political settlement of 1998 (Borooah, 1995; FitzGerald and Morgenroth, 2020). ⁷

3. Institutional geography applied to NI: Explaining and solving the economic predicament

The recent recognition in the literature that there are connections between the UK’s economic weakness and its institutional settlement, as discussed earlier in the paper, and the long-run uneven progress of the NI economy provides clues to thinking about the nature of the region’s economic predicament. It might seem understandable, given the ongoing issues concerning Brexit and the NI Protocol, to interpret the economic problems of NI as largely ones unique to the region. ⁸ However, while there are clearly region-specific factors that help explain its economic problems, many of its problems are far from unique. The UK’s regional economics is far less equal than its peers (McCann, 2021). As has already been noted, economic problems predated the outbreak of civil unrest in the late 1960s. Economic weaknesses on the supply side—such as relatively low productivity and relatedly a small private sector—have likewise persisted long after 1998 (FitzGerald and Morgenroth, 2020).

Following the line of argument presented by economic historians and regional economists, we can categorise the sources of the economic predicament into three categories (Brownlow, 2015, 2016). First, as part of the UK, NI has shared in the long-standing macroeconomic and related uneven regional economic problems of the UK (McCann, 2021; Pabst and Westwood, 2021). The UK, with its allegedly excessive degree of political centralisation, has managed to combine relatively poor productivity performance with a spatial model inferior to many of its peers (McCann, 2021). NI thus in isolation cannot shake off these wider sources of poor economic performance, regardless of its own institutional geography. However, some aspects of the NI economic problem indicate a magnification of wider UK problems. ⁹ This second magnified category might be termed region-specific weaknesses in institutional geography. The third source of problems is however those unique to NI. The ongoing economic complications with the legacy of violence, social division and the Irish border, the last of which helps explain the complexity of Brexit in the NI context, are some obvious contemporary examples (Teague, 2021). This part of the institutional geography can be identified as unique to NI.

A further analytical complexity in understanding the region’s economic predicament is that these sources have ebbed and flowed over time. However, the fact that NI’s economic predicament consists of these three sets of processes, which have ebbed and flowed in their relative importance over time, implies that no single lever can rectify all of these problems. In short, once the region’s economic predicament is understood correctly, there is no single ‘silver bullet’ measure that can put right NI’s many economic issues. This line of argument hence echoes the findings of authors who have considered the historical and contemporary sources of regional economic weaknesses (MacFlynn, 2020; Ó Gráda and O’Rourke, 2022).

4. Institutional geography applied to NI: The limits of ‘economic dividends’

In the ‘first generation’ fiscal federalism literature, the benefits from competition associated with inter-regional migration were supposed to ensure public sector provision and mechanisms best matched voter preferences (Hamlin, 1991; Oates, 1972, 2005; Tiebout, 1956). In this earlier approach, the decentralised systems offered dividends in a number of ways based on alleged greater accountability and capacity for policy innovation (Hamlin, 1991; Rodriguez-Pose and Gill, 2005; Brownlow, 2017). It is notable that the Independent Fiscal Commission for NI in its interim report made reference to empirical studies of these alleged ‘economic dividends’. It concluded that such studies indicated there is an element of risk and reward with fiscal devolution: while it offered risks in terms of forgoing the security of the block grant, it equally offered the rewards of greater efficiency in terms of targeting services to meet the specifics of local needs as well as increased accountability and responsiveness (Independent Fiscal Commission, 2021, pp. 17–18).

To repeat, the institutional geography literature has a different analytical starting point: it suggests that the supposed ‘economic dividends’ of devolution are not automatic; instead they rest on the particular design of the institutions chosen. In this regard, the institutional geography approach is closer to those recent reports that consider UK-wide devolution as it relates to productivity and political economy; these papers paint a less optimistic picture (McCann, 2021; Pabst and Westwood, 2021). While the term institutional geography does not appear in popular discourse, the ‘cash for ash’ scandal, associated with the Renewable Heat Initiative (RHI), brought into the open weaknesses in the NI model of devolution (McBride, 2019). Under RHI, which was designed to be an environmental measure, businesses were to be paid a subsidy to switch from oil and gas to wood pellet-based boilers. However, because the level of the subsidy exceeded the fuel price, RHI actually encouraged excessive boiler use. This largesse combined with a lack of effective cost control ensured that RHI was far from an efficient use of public funds. Furthermore, widespread opportunistic behaviour by a range of actors was far from the world described in ‘first generation’ models. Both investigative journalistic and official outlets, such as the Northern Ireland Audit Office (NIAO), have noted that a large explanatory component of the ‘cash for ash’ scandal was traceable to the way devolution operates rather than merely representing the failings of individual political or administrative agents (McBride, 2019; NIAO, 2020).

In terms of the 44 recommendations in the official report, 11 are traceable to shortcomings found in either the capacity or capability of the Northern Ireland Civil Service (NICS) workforce. A range of organisational problems ranging from project management, commercial ability, leadership and recruitment were identified (NIAO, 2020, p. 74). While the report did not present the findings in terms of institutional geography, we can present the findings as suggesting that the ability of NI to benefit from devolution in terms of gaining ‘economic dividends’ has been constrained by a range of factors including the capacity of the NICS to formulate and implement policies.

Later in this paper we will discuss tax devolution in more detail; it is fairly clear that unless improvements are made to policy administration, the ability to successfully implement tax devolution will be circumscribed. Indeed, we might speculate that far from tax devolution acting in the way that ‘first generation’ models predicted, a weak ability to implement complicated policies may compound rather than eliminate regional economic weaknesses. This line of argument takes us far more into the territory of ‘second generation’ models with their focus on institutional frameworks and the possibility of lobbying. It also suggests that the recent papers by McCann and others concerned with ‘levelling up’, productivity and institutions are more relevant to interpreting institutional geography over the long run.

5. Institutional geography applied to NI: Devolution c. 1921–1972

Of course, devolution in NI began in the 1920s and we can use the concept of institutional geography to help reinterpret the economic aspects of the way devolution operated. In particular, we can see there were efficiency fault lines. Moreover, when we consider the institutional geography of devolution we will see that contemporaries were aware of the design issues that anticipated the Fiscal Federalism literature of the 1950s in that control of expenditure and revenue was not merely randomly assigned (Gibson, 1996; Lawrence, 1965). In the introductory section to this paper, reference was made to the assignment of responsibilities (including financial mechanics) of the GOI 1920. The settlement that was designed in 1920 devolved little in the way of taxation, but it was far from silent on issues of public finance: of the 76 articles and 9 schedules in the Act no less than 14 are concerned with finance (Torrance, 2020, p. 52). Any lack of fiscal devolution granted within GOI 1920 was not due to ignorance on behalf of those who framed the legislation, rather it was a consequence of the centralisation complained about by more contemporary authors (McCann, 2021; Pabst and Westwood, 2021).

Contemporaries were equally well aware of HM Treasury scepticism about granting large degrees of tax devolution to Stormont, and these attitudes found themselves even into textbooks of the time (Lawrence, 1965; Wilson, 1989). As such it reminds us that centralisation, including HM Treasury control of taxation and expenditure, in the UK long predates even the Barnett settlement of the 1970s. For example, while modern Corporation Tax was a product of the 1960s, the authors of GOI 1920 gave thought to granting devolved powers over profits tax to what became the NI Parliament, however eventually they rejected assigning the power to do so (Gibson, 1996). ¹⁰

While this debate over the assignment of taxation may or may not be interpreted as providing evidence of foresight, in other regards the institutional geography did not evolve as predicted. As a number of commentators have noted under the GOI 1920 it was envisioned that the industrial staples of NI in the early 1920s—such as linen, tobacco and shipbuilding—would generate the tax revenues that would pay for its expenditure needs (Johnson, 1985; Lawrence, 1965; Wilson, 1989). In other words, it was intended that NI would have a revenue-led financial settlement. However, the long-run decline of staple industries following on the shock of the Great Depression in the 1930s meant that early on it became clear that such financial settlement could not persist (Johnson, 1985; Lawrence, 1965; Wilson, 1989). As early as the 1938 Simon declaration, which switched the emphasis from revenue to expenditure, in practice NI became an expenditure-led system (Rowthorn, 1981). ¹¹

Another way in which the institutional geography of NI did not work as planned and designed before 1972 was in the area of industrial policy (Brownlow, 2020). The source of this weakness was the strong overlap between the region’s political and business elites when combined with weak controls governing Ministerial conflicts of interest (Brownlow, 2007; Jordan, 2020). The net result was that the relatively generous industrial policy packages on offer, which were tilted towards capital grants and advance factories, tended to be allocated based on political access rather than commercial merit to the extent that half of all subsidies aimed at promoting new industries found itself finding its way into linen production (Brownlow, 2007).

Problems of misallocated subsidies and advance factories were entwined with cronyism, a phenomenon of continued relevance. Contemporary economists saw the close relationship of business and political elites and ease of access to government ministers as offering NI a competitive advantage as a location for inward investment (Brownlow, 2007). However, more recent empirical work suggests that in qualitative and quantitative terms a less optimistic interpretation holds (Brownlow, 2007). It is instead argued that political access to subsidies by the staples slowed the reallocation of resources away from declining industries and this slowed growth.

Furthermore, when in 1963 the incentives were changed in ways that reduced the overlap between business and political elites, the composition of industrial policy-funded ventures shifted towards more efficient (and less politically connected) producers. In short, decades before the ‘cash for ash’ scandal there are reasons to think that ability to promote productivity was undermined by the opportunistic relationship between some political representatives and vested commercial interests. This interpretation of NI under devolution fits better with recent interpretations of UK devolution that emphasise political economy type processes (Pabst and Westwood, 2021). It also suggests that the institutional geography of NI in the area of public finance evolved during the period c. 1921–1972.

6. Institutional geography, policy formulation and implementation: Revisiting Heald’s Funding the Northern Ireland Assembly

In the previous section, we examined how institutional geography shines light on the way we can interpret the historical operation of the ‘Stormont’ system and its economic and financial repercussions. ¹² In this section, we examine an important assessment of devolution 5 years after the signing of the 1998 Agreement and then consider the changes and continuities that are relevant to considering the state of the institutional geography in the 2020s. ¹³ In this section, we want to revisit David Heald’s important (2003) report entitled Funding the Northern Ireland Assembly: Assessing the Options (Heald, 2003). Heald’s report was commissioned by the now defunct Northern Ireland Economic Council (NIEC). In the report, Heald provided an empirical overview of the state of play regarding the empirics of NI’s public finances at the time as well as a survey of the main issues and institutional structures. While the quantitative picture regarding NI has moved on since the overview he provided, it is a testament to the lack of progress in the political and economic political spheres that some of the main fault lines Heald saw as impeding the transformation of economic policy formulation in 2003 remain weaknesses within the region’s institutional geography nearly two decades later.

In the report, Heald after surveying the evidence made policy proposals in a number of areas as it affected the conduct of policy making. Before making his proposals he noted that in the UK constitutional settlement ensured there were severe limits in ensuring credible commitments. Embedding anything in the constitutional settlement was made difficult because the Westminster government could repeal its own primary legislation. Heald suggested that this limitation had been well understood in Scotland in the 1990s when plans had been drawn up for devolution (Heald, 2003, p. 73). Since 2003 this legal situation has changed somewhat as the Belfast/Good Friday Agreement, and the resultant Northern Ireland Act 1998 has an international dimension further developed by the 2006 St Andrews Agreement and the 2018 Buick judgement by the NI Court of Appeal established that NI Ministers—unlike their UK, Scottish or Welsh equivalents—exert full executive authority in their areas of responsibility, within any Executive-agreed programme as endorsed by the Assembly (Northern Ireland Fiscal Council, 2021, p. 17).

Heald in the report makes a lot of specific observations concerning the Barnett formula; these observations will be developed later as they are more germane to the following section where we discuss the institutional geography of fiscal devolution. However, for our purposes in thinking about policy formulation and implementation the four most interesting observations as a counterpoint to the 2020s are in the areas of thinking about the weakness of the ability to formulate and implement new measures (‘policy capability’), the high potential for populism, poor incentives facing civil servants and the absence of mechanisms suggesting new courses of action. The first area in which Heald noted that NI’s ability to formulate policy in 2003 was weak in the area of what he termed ‘policy capacity’ both inside and outside government. Within higher education he noted that public expenditure work would not inevitably emerge from the universities without the right incentive structures (Heald, 2003, p. 77). This low policy capacity he attributed to historical inheritance as problems of small size were exacerbated during the devolved period (when parity provided for matched funding) and direct rule (where imitation of British policy was often the order of the day). Heald noted that this long-standing situation had created a gap that would be hard to fill.

A second area following from the first related to the danger that populism would emerge as a shared ideology at Stormont. The policy capacity aspect was he argued vital to develop because under power sharing, unlike at Westminster, all the major parties had dual roles in both being part of the Executive as well as pursuing their own interests at election time. Heald noted that whatever weaknesses left–right politics offered in other jurisdictions, it at least forced parties to focus on what the state should do, and how it should do it. The danger he noted in a case of mandatory coalition was that absent such a focus that policy formulation would be weak and parties would merely opt for populist posturing, such as ‘spend but not tax’ (Heald, 2003, p. 76). In such a situation, a shared ideology would entrench soft budgets as there would be no political incentive to raise revenues while substantial political returns to spend more. In the NI context, the tendency towards soft budgets has been a recurrent one (Teague, 2021). Any future movements on tax devolution might have to face the fact that any party advocating devolved taxation to pay for public expenditure could place itself at substantial electoral disadvantage.

Two other actors in the situation are also considered by Heald. The civil service administration is presented under direct rule as essentially engaging in imitation rather than innovation in terms of how policies were implemented in the Northern Irish context (Heald, 2003, p. 76). He notes that under direct rule civil servants had weak incentives to develop distinctive policies. Though Heald does not use the terminology, essentially imitation took the form of direct adoption of British policies or minimal adaptation.

The fourth and final aspect highlighted by Heald was that in the absence of higher education having sustained interest in analysing issues of relevance to Northern Irish economic development, a critical mass of work on the economy would not be forthcoming. Heald observed that occasional commissioned reports ‘commissioned on a contractual basis will not resolve this deficiency’ (Heald, 2003, p. 77). Heald instead argued that organisational continuity was required in the area of economic and financial matters. As he observed:

Northern Ireland needs to find a mechanism for promoting long-term research with the potential of inputting into policy development, without becoming too concerned if this sometimes turns out to be ‘off-agenda’ or produces inconvenient results.

(Heald, 2003, p. 77)

From the vantage point of the 2020s, there remains much in Heald’s (2003) assessment that remains relevant in terms of the institutional geography of NI. Moreover, we can compare the ways in which both progress and regress have been made in terms of developing policy. First in terms of policy capacity while there have been some welcome initiatives such as the creation of the Nevin Economic Research Institute (NERI), backed by the trade union movement on an all-Ireland basis, and the creation of Pivotal, the think tank shared between the two Northern Irish universities, as well as the more recent creation of the Northern Ireland Productivity Forum it has not all been favourable. ¹⁴ Both NIERC and NIEC are no more and even the economist’s forum at Invest NI proved to be a short-lived experiment during the 2010s.

In terms of populism both the Northern Ireland Fiscal Council in its more recent report and the Interim Report of the Independent Fiscal Commission, while not presenting their arguments as a response to Heald, make observations that suggest there are a lot of continuities. For instance, the Northern Ireland Fiscal Council in its discussion of the Block Grant and the Barnett formula as it relates to NI notes that soft budget constraints shape the fiscal relationship between the UK government and the NI Executive. In particular, NI’s total share of UK public expenditure is so small relative to the benefits to the taxpayer in NI that there is an inbuilt tendency for public funding to grow relative to revenues (Northern Ireland Fiscal Council, 2021, p. 54). In terms of the revenue side, an observation of the Independent Fiscal Commission is that in terms of both level of contribution and composition, NI differs from the rest of the UK. The report notes that since 1999/2000 average revenue per head has been £1,364 (14.5 per cent) below the UK average and that this figure has increased in recent years, with a near £800 increase in the difference (in nominal terms) since 2012/2013, though there has been a modest decline since 2018/2019 (£124) (Independent Fiscal Commission, 2021, p. 49).

One facet of the debate on devolving taxation in NI is the imprecise term ‘super parity’, a term which covers policy divergence or differentiation. Most of these take the form of reduced costs or charges for businesses or individuals in NI. In other words, there are a number of areas where NI could raise additional revenue or reduce expenditure if policies coincided with other parts of the UK (Independent Fiscal Commission, 2021, p. 68). In the report, the total cost of such measures is estimated to cost between £599 and £690 m of which the absence of domestic water charges (£344.5 m) is much the largest divergence (Independent Fiscal Commission, 2021, p. 69). The existence of ‘super parity’ is an important manifestation of the ongoing relevance of institutional geography type features. ¹⁵

In the area of the civil service behaviour there are also reasons to think that Heald’s comments c. 2003 have ongoing relevance in the 2020s. Again the interim report of the Fiscal Commission notes that an important consideration in thinking about implementing tax devolution is the ability of the NI Civil Service to carry out such policies to the extent that the report recommends that ‘The NI Executive would have to ensure that appropriate structures and people were put in place before any devolution of tax powers’ (Independent Fiscal Commission, 2021, p. 11). This recommendation traces back to recent work by the Northern Ireland Audit Office (NIAO) (Independent Fiscal Commission, 2021, p. 11). The NIAO report mentioned earlier in the paper found that workforce planning, organisational develop and people management were all inadequate (NIAO, 2020, p. 206).

Heald’s point concerning the ‘mechanism for promoting long-term research’ has implications for thinking about public policy formulation in the 2020s. While Heald did not mention ‘Think Tanks’ by name, it was the case that in 2003 Northern Ireland sat outside the ecology of think tanks found in Edinburgh, Dublin or London. ¹⁶ Heald’s observation regarding mechanisms remains relevant, though as of 2022 there has been some improvement in the situation regarding both Northern Ireland’s think tank situation, for instance in the creation of Pivotal, NERI and NIPF. Likewise, the level of attention paid to the devolved fiscal settlement has increased as witnessed by the creation of the NI Fiscal Council and the Independent Fiscal Commission for Northern Ireland.

7. Lessons for Northern Ireland’s fiscal devolution

As noted above, Heald’s (2003) report discussed the public finance situation in terms of the implications of the Barnett formula for the public expenditure and more general economic structure of Northern Ireland. Heald contended that the issue of peace naturally dominated the NI referendum debate in May 1998 and in contrast to the 1997 Scottish or Welsh referenda issues of tax-varying powers were not controversial (Heald, 2003, pp. 31–32). ¹⁷ Others might suggest that it is more plausible, given how little attention is given to public finance in the Agreement’s text, to argue that the NI Parties (as well as the UK and Irish Governments) simply ignored the issue. ¹⁸

As Heald noted, devolution changed the context within which Barnett operated. First, it became a mechanism for transferring resources between tiers of government rather than being a mechanism internal to the UK government. Second, and related to the first feature, the lack of transparency that had previously characterised the formula would no longer be possible (Heald, 2003, p. 32). As Heald suggested, NI relative to Scotland or Wales was a relative late comer to an official overview of fiscal devolution. The ongoing debate from the mid-2000s regarding devolving Corporation Tax (CT) viewed CT variation in isolation from other parts of the revenue and expenditure landscape. It was only as a consequence of the New Decade, New Approach deal in January 2020 that the institutional machinery of public finance and fiscal devolution were finally made part of the political settlement in NI. In the New Decade, New Approach deal there was a commitment to creating an independent fiscal council, which had been earlier envisaged in the Stormont House and Fresh Start agreements, as well as a review of funding models (New Decade New Approach, 2020, p. 41). Duly in 2021 the Northern Ireland (NI) Fiscal Council was established to bring scrutiny to the finances of the NI Executive. Likewise, on the 12 March 2021 the Independent Fiscal Council for Northern Ireland was established.

Both of these bodies offer scope for a degree of transparency and monitoring that was not previously a feature of the institutional geography in NI. At the time of writing (April 2022) both bodies have produced reports. The Northern Ireland Fiscal Council in November 2021 published its overview of the public finances (Northern Ireland Fiscal Council, 2021) and draft on the 2022–2025 Executive Budget (Northern Ireland Fiscal Council, 2022). Likewise, in December 2021 the Independent Commission published its interim report (Independent Fiscal Commission, 2021).

Rather than summarising the contents of these reports in the remainder of this paper, we will consider how the institutional geography lens may be applied to interpreting the reports. Regarding the Northern Ireland Fiscal Council report while it is primarily an empirical survey of aspects of the region’s public finance situation, as noted before there is a recognition that the way devolution is configured in the 2020s may still lead to problems of soft budget constraints found under the older (c. 1921–1972) machinery:

Given the number of occasions on which the UK Government has provided financial support packages to help sustain or restore NI’s power-sharing institutions, it would hardly be surprising if politicians at Stormont were tempted from time to time to exploit this potential source of finance to loosen the ‘budget constraint’ they face.

(Northern Ireland Fiscal Council, 2021, p. 54)

Indeed in the report game theoretic insights are applied to the contemporary settlement and the possible moral hazard problems as they relate to soft budget constraints are highlighted. The possibility that ongoing subsidies may do little to promote productivity and/or may actually promote opportunistic behaviour is noted within the report. In other words, implicit within the survey of the public finances is a recognition of the way the NI institutional geography may do little to rectify the economic problems that give rise to the need for a high fiscal transfer. Indeed, some may go further and argue that soft budget constraints are endemic to the region’s economy (Teague, 2021).

In terms of the proposals regarding tax devolution the interim report suggests that the suitability for devolution is associated with a range of criteria including economic and policy context, legal constraints, accountability, administrative efficiency and economic efficiency. It is noted that these criteria are similar to those found in the 2010 Holtham commission report (Independent Fiscal Commission, 2021, pp. 101–102). While the report does not present these criteria in terms of institutional geography, it is pretty clear that the report does not view tax devolution as offering an unabashed ‘economic dividend’ type argument; the report is clear that no tax can meet all the criteria. Viewed together the criteria in the report provide a way to think about how the design of devolved institutions in the NI context may influence the suitability for tax devolution: in other words, implicit within the arguments are institutional geography lines of argument.

Furthermore, while the report notes that the criteria are similar to that found in Holtham, there is a recognition that the circumstances of Wales in 2010 are not those of NI in 2021 (Independent Fiscal Commission, 2021, p. 101). For example, the report suggests that accountability, because of a NI Executive based on mandatory coalition, matters more in the contemporary NI context. Gratifying in the light of previous work applying the concept of institutional geography to tax devolution in NI is the recognition that the speed and order of implementing reforms—such as devolving taxation—have a measurable likelihood on the success of such reforms after implementation. Hence, design issues previously discussed in the academic literature concerned with sequencing are present in the interim report (Birnie and Brownlow, 2017; Independent Fiscal Commission, 2021, p. 103).

8. Conclusions

In terms of the foregoing discussion, the historical evidence indicates that the NI economy during the period c. 1920–1972 was beset by a number of challenges in which the model of devolution and the relationship of this model to cronyism was just one consideration that undermined productivity (Brownlow, 2007). Moreover, devolution in NI while facing specific issues, such as the legacy of the border, a legacy that has made the Protocol a recent factor in the Brexit debate with important repercussions for both the polity and the economy, has more general implications for the study of the institutional geography of the UK as well as further afield. These more general implications can be summarised as three key points.

Firstly, ongoing issues regarding competitiveness, public finance and the scope for tax devolution within NI should remind us that there are in reality no ‘silver bullets’. The long-standing historical record of devolution tends to confirm the sceptical interpretation of institutional geography rather than more optimistic interpretations of some proponents of fiscal decentralisation. No single policy can solve complicated and layered economic problems. The more general finding is that the supposed ‘economic dividends’ that are supposed to arise from devolution are of relevance. The institutional geography of NI devolution coincided with a location that has had persistent problems with productivity and industrial readjustment. In other words, tax devolution in and of itself will not cure the supply side issues of the region’s economy. Tax devolution should be viewed as an accompaniment to rather than a substitute for attempts at identifying the sources of weak supply side performance. Much less is tax devolution in isolation able to offset deficiencies in human or physical capital investment.

Secondly, when selecting the appropriate taxes to devolve in a jurisdiction, the Northern Irish experience indicates that while we can debate the merits in devolving specific taxes (such as taxes on income or corporate profits), it is the case that we are more likely to see favourable outcomes if combinations of taxes are devolved rather than any one tax in isolation. It is the case that discrete devolved taxes differ in terms of their individual suitability in terms of economic impacts; that is not the entire story as a piece of institutional economics. Once we consider the devolution of taxes in combinations then considerations of how such taxes interact with each other—rather than act or operate in isolation—become germane and/or the appropriate sequence in which taxes should be devolved becomes more relevant. Issues of sequencing are important considerations when we consider real-world tax devolution.

Thirdly, equity as well as efficiency needs to concern us when we think about devolution. Once again institutional geography provides important insights in thinking about the pressures that devolution may give rise to in the area of tax devolution. While we as economists often think about the balance between overall costs and benefits (and/or Kaldor–Hicks type exercises) as providing the way to think about the potential efficiency implications of a public policy proposal, politics as well as economics matters. ¹⁹ As Olson observed long ago it might be the relative concentration or diffusion of costs and benefits that matters in the political ‘marketplace’. ²⁰ In short, egalitarian outcomes may not inevitably flow from devolving taxation even if we regard tax devolution as promoting efficiency. The creation of the institutional machinery of the fiscal council and the tax commission are to be welcomed. What is even more welcome is that the analysis in both reports shows an awareness of institutional geography even if the reports are not explicitly institutionalist exercises.