Levelling up policies and the failure to learn

ABSTRACT UK policy targeting regional economic disparities has been characterised by frequent reversals and announcements, with multiple, uncoordinated public bodies, departments and levels of government responsible for delivery. Prior ‘place-blind' national policies have given way recently to ‘place-based' approaches, with a convergence between industrial and spatial policies. Yet a consequence of inconsistency and poor co-ordination is that the UK policy framework lacks adequate feedback mechanisms from local outcomes to the national policy process; there is a failure to learn. More effective policies to address spatial inequalities require institutional reform embedding evaluation and learning mechanisms into subsequent policy analysis and implementation. Other advanced economies offer institutional examples that could be feasibly implemented in the UK.


Introduction
Recent experience with levelling up policies offers a striking example of deep-seated institutional shortcomings in the UK, particularly with regard to policy churn.In March 2021, mid-pandemic, and to the dismay of many, the Johnson Government abolished the Industrial Strategy launched in 2017 by his predecessor Theresa May, along with the independent Industrial Strategy Council responsible for monitoring it.The 2017 document had seemed to represent a significant shift toward a strategic state-led approach to economic development, with 'Places' as a key strand, for the first time in decades (McCann et al., 2021), a shift which many concerned about spatial inequalities had welcomed.In its place the Johnson Government published a new 'Plan for Growth', which also emphasised the spatial aspect of economic growth: where the 2017 strategy introduced monitoring undertaken by the Industrial Strategy Council, and the introduction of Local Industrial Strategies, the 2021 'Build Back Better' document stated: 'Our most important mission is to unite and level up the country: tackling geographic disparities; supporting struggling towns to regenerate; ensuring every region and nation of the UK has at least one globally competitive city,' (HM Treasury, 2021a, p. 9).
This was not the end of the policy churn.The flagship Levelling Up White Paper was launched in February 2022 by the Department whose name had recently been changed to include the term 'levelling up' (DLUHC, 2022). 1 It created an independent advisory body, the Levelling Up Advisory Council, although this had not published anything in its first 12 months.The short-lived Truss Government then brought out the 'Growth Plan' (HM Treasury September 2022), which did not use the terminology of levelling up but nevertheless featured investment zones around the country, while the successor Sunak Government's statement on 'Stability, Growth and Public Services' in November 2022 focused on levelling up via infrastructure and devolution deals.
A long period of 'place-blind' national economic or industrial strategies has thus given way in recent years to a focus on 'place-based' policies.This was in large part due to the emergence of the 'geography of discontent' (McCann, 2020;McCann & Ortega-Argilés, 2021).Modest devolution within England has occurred with the introduction of city deals and directly-elected Mayors in different places; but their economic powers are limited, although modestly extended for some places in the 2023 Budget.Most placebased policies and funding are determined centrally, and cascaded through a 'fragmented system of overlapping authorities at the sub-national level,' (Richards et al., 2023, p. 1).At local level individual organisations are accountable to different central authorities, reproducing central silos thanks to the absence of an overarching authority (Solar & Smith, 2022), while the abolition of the Audit Commission in 2016 adversely affected the ability to monitor progress (Redmond, 2020).The levers available to devolved English authorities seem insufficient to drive improvements in productivity and growth (Tilley et al., 2023).
Each of the recent economic policy documents acknowledged, to a greater or lesser degree, the need to address spatial inequalities, yet the national policy delivery framework and mechanisms changed four times in five years.This churn was not an aberration in British economic policy-making, but symptomatic of the underlying institutional context (Martin et al., 2022).The decline of institutional policy memory in the UK has been previously noted (e.g.Corbett et al., 2020;Pollitt, 2000Pollitt, , 2009;;Stark & Head, 2019), despite the introduction of some entities designed to preserve knowledge, such as the What Works Centres (including one for Local Economic Growth).
The recent UK policy instability can be juxtaposed against the fact that policymakers in other OECD countries have been expressing over the same period renewed interest in more strategic, developmental approaches to the economy (Aiginger & Rodrik, 2020).The British economy has experienced low growth and poor productivity performance since 2008, and regional inequalities are extreme by comparison with other OECD countries (Carrascal-Incera et al., 2020;McCann, 2020).These point to underlying structural weaknesses in the way the economy is managed, particularly the interaction between policy churn and incoherence in national decision-making and delivery as it affects outcomes in different parts of the country (Bailey & Driffield, 2007;Richards et al., 2023).
This paper argues that to deliver on the ambition (and need) to level up, the UK must improve the efficacy of its policy process and the institutional arrangements to deliver successful interventions.The policy churn, policy silos and fragmented and often incoherent spatial governance structures characterising the UK combine with the absence of institutional mechanisms for learning from experience.Other countries offer a number of potential models for institutional learning.
Based on an analysis of the key industrial policy documents produced by the British government since the 1960s, we map the top-down institutional arrangements and relationships that constitute the economic policy environment for local economic outcomes, arguing that there is a fundamental institutional weakness in attempts to pursue a Levelling Up agenda due to the system's inability to learn from local outcomes.In comparison with other countries the lessons drawn for local outcomes are not institutionally linked back to national policy decisions, nor to local policy implementation through the fragmented and overlapping national and sub-national authorities concerned.From the mapping of the process of generating and implementing policies, we identify key deficiencies.These are: a lack of information feedback from economic evaluation in specific places to national policy-setting, hence ignoring local knowledge in policy-setting; and, as a result of lessons not being captured institutionally, policy inconsistency and co-ordination failures.We compare the UK with some other advanced economies and consider ways the UK could rectify its institutional failings, by embedding a learning mechanism and better co-ordinating policies.
We begin by describing the way the climate of ideas concerning the role of the state in spatial economic development has shifted back in favour of strategic intervention, describing briefly some past successes and highlighting the benefits of policy consistency.We then (in Section 3) contrast the UK's policy churn and lack of co-ordination with examples of strategic policymaking elsewhere, highlighting the lack of any institutional pathways for learning from experience in the UK.We discuss (Section 4) some examples of institutional models that embed learning, concluding that institutional reform in the UK would mitigating the problems of lack of consistency and lack of co-ordination in levelling up policies.

The return of strategic government intervention
For decades, views about economic policy have polarised around the respective roles of market and state.Although there have been many microeconomic or sectoral economic policy interventions, as documented below, governments in the UK since 1979 have expressed a particular aversion to active 'industrial policy' as a vehicle for strategic coordination of economic development activity by the state (Chang et al., 2013;Stiglitz & Greenwald, 2014;McCann et al., 2021).In the context of developed economies, industrial policies are viewed by advocates as a purposive way of shaping the structure of production in the economy toward higher value activities and to benefit important political geographies (Andreoni et al., 2019).
The prior record of such strategies differs greatly between economies, with some having longstanding economic strategies.Notable success stories include the East Asian 'miracle' economies, first Japan, later others including South Korea, Taiwan and Singapore.Their whole-of-government frameworks were explicit attempts to create a domestic production structure that was globally competitive.Many of today's well-known Asian-based multinational giants, such as Toyota, Mitsubishi, Samsung, Hyundai and Taiwan Semiconductor Manufacturing Company (TSMC), can trace their growth to purposive policy measures. 2Such measures include encouragement for the formation of industrial conglomerate networks such as South Korea's chaebols and Japan's keiretsu (Studwell, 2013).Although it has experienced slow growth since the 1990s, postwar Japan is often cited as a textbook example of a well-coordinated whole-of-government (national and local) economic strategy, featuring strong collaboration between ministerial departments and agencies.
The American experience of long-term strategic intervention is often overlooked, but there is renewed recognition of the role government policies there have played in driving long-term development and structurally transforming the economy.Measures have included subsidies for strategic industries, tariff rebates on imported inputs, and use of technical standards to reduce uncertainties for investors (Chang & Andreoni, 2020;DeLong & Cohen, 2016;Wade, 2014).The US government continues to support key sectors, such as defence and energy, through channels ranging from basic research funding provided by bodies such as the National Institutes of Health (NIH) and the Advanced Research Projects (ARPA) agencies 3 to dedicated subsidies for private enterprises like SpaceX and Boeing, and procurement programmes.
Historical accounts of Britain's experience identify many policy mishaps in the 1960s and 1970s (for example, Crafts &Hughes, 2013 andCrafts, 2018).Nationalisation and subsidies were the primary tools in the postwar years, used to support declining firms such as British Steel Corporation and British Leyland.Attempts to create new economic engines, such as commercial nuclear power, did not lead to sustained success. 4These experiences left a generation of British policymakers with lasting distaste for 'picking winners', reinforced from 1979 by an explicit ideological push to reduce the role of the state.This Thatcherite political narrative linked state intervention to Britain's economic decline, so subsequently the received wisdom in UK policy circles has been that the best government economic strategy is to leave growth to the market (Bailey et al., 2023;Egerton, 2018).The impact of the early 1980s recession on UK manufacturing brought about a shift toward financial services, further enabled by the Big Bang deregulation of 1986, the major privatisation programme through the London stock market, and support through infrastructure investment in London's Docklands (Davis, 2023).This shift to financialisation and a knowledge-based or even rentier economy amplified the spatial inequalities already emerging through deindustrialisation: London thrived while the industrial areas of the UK declined (Christophers, 2019;Martin et al., 2022).What's more, high value manufacturing activities have also shifted away from the traditional industrial regions (Sunley et al., 2021).
A broad shift in economic philosophy toward strategic government intervention has occurred in many developed economies as a result of recent economic shocks.The EU has updated its Industrial Strategy, supporting emerging green and digital transition industries such as hydrogen energy, cloud computing and electric vehicles (BMWi, 2021;European Commission, 2021a).Across the Atlantic the Biden Administration has recently significantly expanded the state's role through a multi-billion dollar public investment programme in innovation (in areas such as green tech) and key industries (such as IT) through five pieces of legislation: the US Innovation and Competition Act (USICA), the American Rescue Plan Act, the Infrastructure Investment and Jobs Act, the CHIPS and Science Act, and the Inflation Reduction Act (congress.gov).
At the same time, newer approaches to economic policy are emerging in the academic literature, synthesising classic market failure arguments with new approaches to learning over time and the accumulation of ideas in the context of endogenous growth (Stiglitz & Greenwald, 2014).The rationales for intervention in this recent literature can be summarised as compensating for information failures, co-ordinating private sector actors and de-risking investment in the context of technological transitions (Cherif & Hasanov, 2019;Noman & Stiglitz, 2016;Rodrik, 2019b).In the UK context of levelling up, we therefore focus on the information feedback mechanism within policy-making, looking at what types of information and learning can occur during the policy generation and implementation process.
The UK's current economic policy discourse stands out in terms of its languagethe levelling up terminology is somewhat distinctivebut not in terms of its shift toward a more active developmental role for the state.The persistence of spatial inequalities cemented by the 1980s experience of deindustrialisation for major UK cities outside London has gained increased salience thanks to the 'geography of discontent' (McCann, 2020) reflected in the 2016 Brexit vote and 2019 'red wall' voting phenomenon (McCann & Ortega-Argilés, 2021).There is therefore growing policy interest in place-based industrial strategies, in a revival of interest in both regional policy and industrial strategy, and their overlap (Bailey et al., 2023).

Weaknesses of the UK policy-making context
In the context of this re-emerging interest in strategic economic development, the weaknesses in the UK that stand in the way of successful levelling up can be divided into two categories; lack of consistency and lack of co-ordination.Given the centralised UK model, it is largely national policies that set the governance framework and opportunities for local or regional decisions and economic growth.Incoherence at the centre therefore has substantial implications for the spatial distribution of economic activity.
The lack of consistency, or policy churn, has been widely noted.Policy-making practice is generally described as the net result of two competing styles of policy-crafting processes.The first is an 'impositional' (top-down) style, where policy proposals originate at the political level, i.e. ministers and their staff, while the second reflects a 'consensual' (bottom-up) style of policy-making, emphasising collaborative policy process via networks of affected interests and relevant civil servants (Richardson, 2018).Richardson notes that in most aspects of government the UK has gravitated towards the impositional style.This, at its worst, is striking for its dearth of incorporation of insights from important economic actors with local knowledge, giving rise to what Richardson calls 'pop-up' policy-making.
The impositional style of policy-making leads to numerous policy changes with no improvement in the Government's capacity to solve problems effectively (Richardson, 2018).The short-term nature of policy-making leads to several problems.The first casualty tends to be the interruption of the structural transformation the local policies were intended to catalyse.There is frequent upheaval in contexts where the private sector has to make long-term decisions, such as education or infrastructure (Institute for Government, 2017).Structural reforms such as those needed for levelling up to happen can only occur over longer time horizons.The lack of consistency in policies over time also makes it difficult to evaluate their impacts and cultivate long-term institutional capability.In short, there is no capacity in the system to learn.
It is perhaps not surprising therefore to see that the British approach to levelling up has been ad hoc and haphazard, characterised by regular cycles of new policy announcements that are often rolled back (Norris & Adam, 2017).The British approach to microeconomic policies in general has been linked to political cycles; episodes of policy change reflect the prevailing paradigm of the day (Norris & Adam, 2017).Figure 1 illustrates just a few of the policy changes in the national economic frameworks concerning major regional or local policy initiatives.
As noted in the introduction, the post-2017 period of levelling up policy amply demonstrates the churn.The abolition of the 2017 Industrial Strategy did not involve any consultation, including with the many local authorities developing Local Industrial Strategies at the time.Supplementing the 2021 Plan for Growth and Innovation Strategy were proposals to create several new government bodies, including the Advanced Research & Invention Agency (ARIA), modelled after the US's Defense Advanced Research Projects Agency (DARPA), with £800 million of public funds to spearhead transformative R&D programmes (Department for Business, Energy and Industrial Strategy (BEIS), 2021a).The new UK Infrastructure Bank, to finance green and regional infrastructure projects, has a projected financial capacity of £22 billion (HM Treasury, 2021b).A National Audit Office evaluation stated: 'The UK Infrastructure Bank was set up quickly, and there is more work to do before it is fully operational and is able to support the government's aims in achieving net zero and supporting local economic growth' (NAO website, 2022).
Some of the specifics in the suite of policies post-2017 do represent continuity, such as a consistent focus on AI.However, aspects of levelling up policy have been left in limbo (such as local industrial policies) and the status of many microeconomic levelling up policies remains unclear.Moreover, the Plan for Growth, the Innovation Strategy and the subsequent 2022 Levelling Up White Paper and the 2022 Growth Plan, lack any mechanisms to independently track, evaluate, andif necessaryrecommend changes to existing strategies and policies.A particularly striking illustration of lack of consistency in the levelling up context is provided by the example of the shift from a regional focus and the 1998 creation of the Regional Development Agencies under New Labour, and their 2010 abolition and replacement by under-resourced and geographically fragmented Local Enterprise Partnerships (LEPs) (Fai et al., 2022, pp. 746, 753).LEPs were intended to shift the focus to local growth and business voice (Hildreth & Bailey, 2013, pp. 237-238).However, their staffing and activities overlapped with local authorities and there have been several reviews of how they functioned, suggestive of dissatistfaction: in 2017 (DCLG, 2017), 2020 (BEIS, 2020) and again in 2022 (DLUHC, 2022), before the March 2023 Budget announced a consultation on their abolition, with a view to handing their (limited) powers and resources to 'local government' (HM Treasury, 2023, p. 3.119).
In addition to policy churn, there is a lack of policy co-ordination, at the national level and also between national and local decision-making and accountability.Since policy is largely generated by a 'top-down' approach by ministerial departments, the institutional structure for collaboration and co-ordination is inherently absent, and depends on buy-in across senior politicians.Lack of cross-government co-ordination is frequently cited as an impediment to effective policy rollout (BEIS Committee, 2021).Cross-ministerial functions remain underdeveloped in the UK, only occasionally being assumed at various times by key powerbrokers in the Treasury, the Cabinet Office or the Prime Minister's Office.In addition to the departmental silo problem, the position taken by the Treasury, widely regarded as the most powerful institution in British policy-making, usually dictates the outcome of policy deliberations.The Treasury's focus on other important policies such as fiscal management often means that supply side policy often gets a lower priority, and also falls foul of the Treasury's institutional aversion to intervention by government.
The fragmentation of policy delivery extends to the agencies housed in various departments, including BEIS and HM Treasury.This includes institutions that engage frequently with local businesses such as the UK Export Finance (UKEF), the British Business Bank (BBB), British Patient Capital (BPC), or the Start-Up Loans Companynone of which have any explicit programmes relating to levelling up (or economic strategy more broadly) in their past or future business plans, despite its prominence in national policy debate (BBB, 2020;BPC, 2020;UKEF, 2020).Moreover, the support measures from the BBB and other agencies are themselves formulated and enacted by top-down decree rather than generated bottom-up based on conditions and opportunities discovered by the local programme managers.
Given that senior ministers, their teams, and ministerial departments are afforded significant freedom in dictating the overall shape of policy, 5 it is unclear how they weigh the relative importance of evidence and analysis as opposed to political calculus and lobbying.Political turbulence has contributed to the state of flux.Nevertheless, there are several key organisations that have a either a continuing influential role in shaping future policies, or specialised roles in delivering policies.Figure 2 presents a simplified landscape of the institutions for microeconomic economic policy analysis and implementation.These include of course HM Treasury.In addition to being the key department designing overarching economic strategy, the Treasury also houses the UK Government Investments (UKGI), the primary holding body for many key state-owned enterprises.Institutions such as the BBB, the parent company for agencies like the BPC and the Start-Up Loans Company, are managed by UKGI despite being nominally owned by BEIS (NAO, 2020).Another executive agency of the Treasury, providing independent advice on major public infrastructure decisions, is the National Infrastructure Commission, which has undertaken a number of studies of the impact of infrastructure on spatial aspects of growth.Coupled with the fact that the Treasury is the final approval body for expenditure, it has unparalleled influence over all aspects of policy, including matters of local growth.
Nevertheless, the co-ordination challenge when it comes to levelling up is daunting.The Department for Levelling Up, Housing and Communities is primarily responsible for delivery of levelling up policies and local industrial strategies, and for local government.Other departments have responsibilities with spatially-relevant consequences.The February 2023 break-up of the former Department for Business, Energy and Industrial Strategy (BEIS) saw the formation of three new ministries: the Department for Business and Trade (DBT), the Department of Science, Innovation and Technology (DSIT), and a new energy and net zero Department.The new DSIT published a new Science and Technology Framework in March 2023; it houses UK Research and Innovation (UKRI), the primary source of grant-related funding of research, whose uneven geographical spread has come into focus (NAO, 2020;Forth & Jones, 2020).The Department for Transport is also key ministry for transport decisions affecting the spatial distribution of economic activity, in particular the appraisal of transport projects, but without a levelling up remit (Coyle & Sensier, 2020).As well as ministerial departments the UK has (at present) 425 public agencies and other bodies reporting to various departments (gov.uk).Some have responsibilities that can significant spatial economic consequences, such as the funding decisions of the Arts Council England or minimum wage recommendations of the Low Pay Commission.
While Figure 1 illustrates frequent national policy changes affecting regional and local outcomes with the political cycle and Figure 2 demonstrates that the national landscape for locally-salient policies is institutionally complex, Figure 3 illustrates one example of a further complexity, the geographically overlapping and multiple tiers of local government jurisdictions in many parts of England.The lack of coherence varies across places.In this case, East Anglia is illustrated where the complexity is particularly acute.In some others, such as the uniquely coherent Greater Manchester Combined Authority (illustrated in Figure 4 for comparison), and in the devolved nations, there is much greater coherence.Elsewhere there is considerable overlap and lack of clarity about accountability and indeed even data collection for the purposes of evaluationin contrast to other countries with more 'general purpose' administrative geographies (Newman & Kenny, 2023).
Together, these three features illustrated in Figures 1-3 a lack of consistency, and a lack of co-ordination, at national level in spatially-relevant policies, between national and local institutions/policies, and between different policy geographies and tiers of governmentunderline the challenge inherent in embedding the ability to learn in an effective levelling up or spatial economic development strategy.
There has been a considerable amount of academic and government evaluation of policy impact, including by the What Works Centre for Local Economic Growth.The What Works Centres are tasked with collating evidence on the effectiveness of interventions in their respective areas and presenting findings accessibly.The Centre for Local Economic Growth has produced impact evidence reviews on subjects ranging from estate renewal to the impact of fast broadband.However, a subsequent mechanism for learning from these and other evaluations, and subsequently implementing lessons, is not institutionally embedded into the UK policy process.The practice of incorporating consultations from expert groups and relevant local interests is also haphazard, with little clarity as to how the information derived should be incorporated into the policymaking process.This is a particular issue for levelling-up policies given the inevitable information gaps at the centre about local economic conditions.There is no channel for independent evaluation of policies with the intention of informing future policy decisions.This vacuum created by an absence of consensual evidence-based statecraft leaves space for 'ideational' policy proposals to take root, ones driven by political actors and stemming from political motivations (Richardson, 2018).Ministers and their respective teams control the process of policy generation.The complex map of local government also limits the scope for effective learning from local know-how.
While the political process in a developed Western democracy such as the UK does, to an extent, act as the medium through which dispersed sources of information around the country are discovered and incorporated into policy, relying on this political channel alone is likely to be suboptimal.In complex domains such as supply-side economic policieswhere the time horizons of structural transformations are long and uncertain, while gains are diffuse and difficult to quantifyheuristics and biases in political decision-making militate against evidence-based judgements (Steenbergen & Colombo, 2018;Vis, 2018;Wade, 2014).This missing element of learning locks the UK into shortlived, politically-driven policies.Its absence leads to the two particular deformations of policy in the UK we have identified: lack of consistency and lack of co-ordination.The UK model stands in contrast with examples of international practice involving rigorous oversight and evaluation by independent and arms-length bodies, and feedback into subsequent policy development.A consequence of the UK's failure to embed evaluation results in subsequent analysis and implementation is that policy discussions become strongly skewed towards tangible short-term costs with less emphasis on potential gains, which are typically diffuse, long term and hard to quantify.One example which had significant spatial implications was 3i Group (formerly Industrial and Commercial Finance Corporation (ICFC) and then Finance Corporation for Industry) being divested from the Bank of England's holdings in 1987 and publicly listed in 1994.This transformed 3i from a key funder of SMEs across the UK into a nationally focussed private equity firm, leading to a long-term decline in SME financing (Collier & Mayer, 2020).The perception of a resulting lack of SME and start-up finance led to the creation of a new policy vehicle in the form of the BBB in 2014, and then BPC in 2018. 6In contrast, the German state-owned investment and development bank, KfW, works through the local banking system and has been in continuous operation since 1948 (Haldane, 2018).Similarly, the Japan Development Bank (JDB, or Development Bank of Japan today), with 10 regional branch offices and eight representative offices around the country, has largely retained its central policy role of economic development since 1951 (Shimada, 2016).Such examples feature a prominent institutional role for 'bottom up' processes and local knowledge, and link economic strategies to place-based outcomes.The next section turns to some examples of how learning is institutionally embedded in the policy process in Japan, the US and elsewhere.

Learning institutions and mechanisms
Anchoring Japan's postwar strategy was an institutionalised bottom-up process for generating policies.Japanese technology and sectoral policies have an integral regional dimension (Bass, 1998).Reducing regional inequalities was central to Japan's postwar economic development, leading to its becoming one of the least regionally unequal OECD countries.As part of the suite of policies, which also include transport policies such as high-speed rail investment (Miwa et al., 2022), MITI's junior officials gather information from local and industry stakeholders (Ohno, 2018;Figure 5).This influences subsequent policy recommendations through an explicit institutional mechanism (Figure 4).There is also co-ordination across government, including the Fiscal Investment Loan Program (FILP), 7 administered by the JDB (an independent agency under the Ministry of Finance) (Ohno, 2018;Shimada, 2016) and the nationwide network of Shindanshi (state-certified SME management consultants).
In the US, less often thought of as having a purposive industrial policy, policy success in creating one of the dynamic, innovative and knowledge-intensive economies in the world has been attributed (e.g.Liu, 2020;Van Reenen et al., 2021) to a similar model of 'bottomup' decentralised approach, spearheaded by institutions such as the ARPA agencies, the National Science Foundation (NSF), the National Institutes of Health (NIH) and extensive networks across industries, academic institutions and other government agencies.These institutions avoid political gridlock thanks in part to their sensitivity to the country's economic geography (Wade, 2014).The emergence of Silicon Valley as an innovative, researchintensive cluster, for example, was helpfully seeded by allocations of defense procurement and research funds specifically to California (O'Mara, 2020).The Biden Administration's recent large-scale industrial policy approaches have an overtly placebased element (Parilla et al., 2022).On an operational level, US policy connects actors across innovation ecosystem, through agencies such as DARPA (Fuchs, 2019).
In contrast to such regionally-sensitive approaches, the UK's policy practice implicitly assumes that central government has near perfect information and visibility of local conditions.This mode of policy-making presumes that market failures can be identified, that the centre knows which to prioritise, and what types of policy interventions are required (Fernández-Arias et al., 2020).In reality, relevant information on where market or government failures, obstacles to structural changes and barriers to investment exist may be diffused widely across the economy (Fernández-Arias et al., 2020;Rodrik, 2019aRodrik, , 2019b;;Romer & Griliches, 1993).
Better policy processes require an institutional framework that would enable the following: . An effective capability for local information gathering and analysis; . An institutional means of bringing together dispersed information; . A learning mechanism and accountability mechanisms to ensure that information gathered and evaluation of outcomes shapes subsequent levelling up policy design implementation.
There are several possible approaches to establishing such frameworks.One is independent scrutinysuch as that provided by the former Industrial Strategy Commission on levelling up aims (as opposed to the advisory role of the more recent Levelling Up Council).In other policy domains, such as fiscal policy or climate policy, the UK has independent scrutiny bodies.Experiences from these suggest improvements in the quality of subsequent policy decisions. 8In the fiscal policy domain for example, the indirect influence of independent institutions like the UK's Office for Budget Responsibility (OBR) or the US's Congressional Budget Office (CBO) on the policy-making processvia pre-emptive effects on budget preparations, informing public debate and fuelling political actionhighlights their capacity to shape policy.
Importantly, a body of this kind can also provide a co-ordination function.Examples overseas include the Netherlands.Its Bureau for Economic Policy Analysis (Centraal Planbureau, CPB) in the Netherlands conducts independent evaluation of policy proposals.Particularly since the 2004 Spatial Memorandum (Nota Ruimte), CPB evaluations have where relevant included a place-based element, but evaluate outcomes across functions (Bos, 2010).Regional development policies in South Korea have taken various forms aimed at reducing regional disparities; the Korea Development Institute (KDI) has undertaken critiques of policy across the board (e.g.Kim, 2008).The legal remit of the Australian Productivity Commission includes promoting regional development as one of it eight mandates, while it too can take a holistic view of how government policies contribute to specified aims.
Another type of co-ordinating independent institution of which examples are found overseas are sovereign wealth funds, also called sovereign developmental or venture funds, exemplified by Singapore's Temasek Holdings, Australia's Future Fund, or Malaysia's Khazanah Nasional (Alsweilem et al., 2015;Dixon & Monk, 2017).Here, the state typically utilises independent investment institutions to implement policy aims alongside competitive market forces to ensure financial discipline (Bruce-Clark & Monk, 2017).The Future Fund comprises funds each of which has an investment mandate based on legislation.The Singapore and Malaysian funds are more opaque, answering to Boards of Directors which may be vulnerable to political capture.Each could be considered as akin to private equity firms but with a long-term public service remit (Bruce-Clark & Monk, 2017;Santiso, 2017).What distinguishes these institutions from a standard fund is that they contain dual public wealth-management and developmental mandates.Other similar institutions include Sixth Swedish National Pension Fund (AP6), the Korea Investment Corporation and the United Arab Emirates' Mubadala Investment Company.While these are national institutions, their remit includes a regional or spatial aspect, such as the Khazanah Nasional's 'building vibrant communities', or a requirement to serve the whole of the nation.
Some authors have advocated a new institution for the UK (Detter et al., 2020).But there is no reason an existing body such as UK Government Investments could not be transformed into a developmental levelling up investment institution, given an appropriate mandate and governance.Alternatively, state investment banks could be a vehicle for levelling up policies.Indeed the British Business Bank has explicitly asked for greater independence in order to be able to act like a sovereign growth fund able to make long-term decisions (Thomas, 2023).In either case, the mandate could explicitly include co-ordination among other bodies in the fragmented spatial institutional landscape.Any of these forms of independent institution receiving bottom-up local information, and co-ordinating across the many national and local bodies, could deliver on the need for local information-gathering and the bringing together of dispersed know-how.
The final missing element in the UK is the institutional link from learning and co-ordination of information back to the policy-making process.Existing evaluation bodies such as the What Works Centre for Local Growth cannot achieve this, nor do elected mayors and Combined Authorities have sufficient powers or leverage to shape national policies that will affect them.Recognising the centrality of learning in driving structural economic change over time, the options discussed heresuch as oversight agencies or development funds with institutional longevity and a local information-gathering rolerepresent potential ways of embedding learning into the policy generation process and achieving more effective levelling up.In the absence of further political devolution placing more levers in the hands of sub-national government, independent institutions seeded by existing bodies in the UK landscape and given a specific mandate of regional economic development could gather information, co-ordinate the many actors involved, and would have some levers to advance levelling up.

Conclusion
The central argument of this paper is that inconsistent and poorly co-ordinated policies for levelling up in the UK stem from its institutional structures.Inconsistency arises from an 'impositional' policymaking style where local long-term structural transformations are interrupted by frequent changes in national policy.Uncoordinated policy arises both from the fragmentation of policymaking bodies, including national bodies detached from local stakeholders that do not explicitly prioritise levelling up, and also CONTEMPORARY SOCIAL SCIENCE complex overlapping systems of local government.Fundamental to both problems is the absence of mechanisms for feedback and learning from local outcomes, in contrast to overseas examples of such institutionally embedded channels.
This has led to a constant cycle of policy churn with no integration of delivery mechanisms.Indeed, the most recent period in British politics has been unusually unstable, and the weaknesses of national-to-local policy-making have been cemented by the abolition of institutions such as the Audit Commission, a particularly important 'learning' institution in the levelling up context.Therefore, in addition to improving and making systematic information gathering, analysis and collation, the paper recommends improving learning and accountability mechanismssuch as by establishing independent oversight agencies or development fundsto ensure that information gathered and evaluation of outcomes shapes subsequent levelling up policy design and implementation.ARPA-Energy (ARPA-E), and Homeland Security ARPA (HSARPA).4. Britain's strategy to commercialise nuclear power in 1970s had several critical missteps (Coyle, 2020): The first was choosing a reactor type (Advanced Gas-cooled Reactor, or AGRs) that no other country used, which meant that Britain could not take advantage of exporting supplies or know-how to other markets.The second was inconsistent signals of public sector support to the marketthe commissioning of two more AGRs in 1978, before backtracking and switching to light water reactors (widely used in the US and elsewhere) created immense technological uncertainty in the industry.5.It is well-established that the UK's Westminster model of majoritarian democracy confers a significant degree of latitude for the ruling party to make nearly unilateral executive decisions (Liphart, 2012).6.The BPC, which has already played a significant role in deepening the venture capital pool in the UK, is already set to be privatised in the foreseeable future.This stands in contrast with many other similar patient capital investment vehicles in other nations, such as Israel's Yozma or New Zealand Growth Capital Partners (NZGCP), which continues to be retained as a key policy delivery vehicles to assist in deepening venture capital pools as new technologies and industries continues to emerge in the future.7. FILP was a mechanism in which funds from postal savings and pension contributions from the private sector were mobilised to conduct investment and loans having public nature (typically infrastructure construction and business support) through state institutions and credit mechanisms.Its financial resource was at times as large as half of the central government's general budget.8.For example, see Averchenkova et al. (2018) for an overview of how the Climate Change Committee has changed and improved the climate policy-making practice in the UK.

Figure 1 .
Figure 1.Simplified timeline of major UK industrial and regional economic policy developments.Adapted by the authors building on Norris and Adam (2017), IfG (2017).

Figure 2 .
Figure 2. Simplified map of key British economic institutions, initiatives and their relationships.

Figure 5 .
Figure 5.Japanese industrial policy generation process.Adapted by the authors from Ohno (2018).

Notes 1 .
The Department for Local Government (established by the then Labour Government in 2006) was renamed the Ministry for Housing, Communities and Local Government in January 2018, and the Department for Levelling Up, Housing and Communities in September 2021.2. For instance, Shih and Wang (2010) documented Taiwan's industrial policy journey to create a sophisticated, technologically-advanced, and capital-intensive economy.3. Includes Defense ARPA (or DARPA), as well as other ARPAs, such as Intelligence ARPA (IARPA),