Addressing macroeconomic stability and regional inequality in the 21st Century

A proposal

Introduction and (relatively) non-technical summary

The core idea of this proposal is that macroeconomic stability is good for growth but this must be shared by all in order to achieve quality growth. That is not to say that macroeconomic stability by itself is the be-all and end-all but that it makes all other policies either easier to achieve or possible at all.

Therefore this paper proposes altering the mandate of the Bank of England to incorporate the lessons of not only the financial crash and great recession but also the economic lessons learned over the last century. This introduction will specify three, tightly controlled and targeted, new tools that would be given to the bank but also the reasons why they are appropriate to the situation the UK currently finds itself in.

The UK is currently experiencing low growth generally but what is increasingly being noted is that there are large regional and individual differences in economic growth, with inequality between individuals remaining an issue but regional inequality is a large concern as well. These regional disparities follow not only the traditional north-south divide but also a rural-urban divide and a cities vs towns divide with towns falling noticeably behind.

The second point worth noting is the many proposals for investment in infrastructure, green technologies and/or a new industrial strategy. The IPPR has focused on this approach recently. So why not write a paper on that instead? It’s not because this paper stands in opposition to those proposals, quite the reverse, it is intended to facilitate them. It may be economically rational to borrow to invest when interest rates are low and the economy is stagnant but it has proved politically extremely difficult. By contrast during the New Labour years it was possible and popular for the government to invest in new schools and hospitals, radically improving lives in the process. Macroeconomic stability facilitates sensible governance, while instability opens the door for reactionary Conservatives and authoritarian Nationalists alike.

The worries that made austerity seemingly politically mandatory in 2010 centred around the national debt and how it would be paid for. While austerity may have been the political consensus then it seems to now be broadly agreed in academic economics circles that it did slow UK growth. Any proposal for a new macroeconomic stability settlement should attempt to deal with both so-called “deficit bias” (which it seems to be acknowledged that New Labour was guilty of, if only quite mildly so) and austerity bias, as we saw during the post-2010 period. While also dealing with concerns around the national debt.

Essentially we must counter the political bias towards pro-cyclical economic policy that is ultimately damaging.

We should acknowledge both that Keynesian insights into aggregate demand really do matter, as the painful failure of austerity shows, but so does the importance of maintaining low and steady inflation, preferably with low (real and nominal) interest rates.

The previous model for solving this issue was an independent central bank controlling short-run interest rates in such a way as to target low but steady inflation, usually interpreted to mean roughly 2% (or just below 2%) consumer price inflation. This was successful in some respects as low inflation was achieved, and although this did create some extra unemployment, the Nordic model shows it is possible to run this system and still achieve reasonable social justice via the inclusion of a strong welfare state. However the model falls down in three respects: concerns that it forces the creation of excessive and unpayable debts (this is damaging if true and also damaging if it is merely believed to be true as it will affect political and economic decision making), that central banks are unable to fulfil their mandate once they hit the zero lower bound on interest rates (even with unconventional monetary policy such as quantitative easing) and finally that the interest rate is uniform across the monetary area and therefore ignores divergent regional economic needs.

The proposal laid out here is this: if politics has an inherent bias towards pro-cyclical fiscal policy and central banks can “run out of firepower” at the zero lower bound then we need to transfer a small number of limited fiscal powers to central banks. The concept is not that this should undo the democratic will, instead it suggests that all political actors want macroeconomic stability but short-run political pressures make this difficult in extreme economic environments and so it is beneficial for the aims of all political actors if an independent central bank can provide that stability. This was the logical behind independent central banks setting interest rates and that logic is continued here.

The reasoning behind each of the three powers listed here will be expanded in the following section. The proposed powers are:

1. A universal citizen’s dividend payable to all citizens that increases at just above the inflation target by default, but this increase can be adjusted if other tools fail to curb excessive inflation or underinflation.
2. A land value tax altered based on a land value inflation target set below the consumer price inflation target but above 0%.
3. A central bank issued bond with no maturity date, effectively an annuity, where core banks (those with a bank account with the BoE and therefore able to hold digital reserves) would have a reserve requirement denominated in these annuities and the reserve requirement would be altered to increase or decrease aggregate demand for these annuities such that their market price causes the nominal yield to equal the inflation target on average.

The proposal in more detail

A running problem for the current arrangement is that the “natural rate” of inflation is unknown and unmeasurable and so a lot of guesswork is required on the part of the monetary authorities. These new tools are all clearly targeted and can be externally and clearly measured. While inflation is still used, it is hard to measure precisely and so it is used as a backstop measurement rather than the first port of call.

Maintaining aggregate demand is a vital task for macroeconomic stability and having moved away from activist fiscal policy to maintain full employment, as in the postwar consensus, the current macroeconomic assignment relies on more generous credit conditions increasing borrowing and filtering down to consumers. The model proposed here uses a hybrid approach- an automatic and oversized fiscal stimulus, effectively permanent helicopter money, via the universal divided rising above inflation, that is tempered by lesser or greater taxation via the variable land value tax and loosening or tightening of credit conditions. If all else fails the citizen’s dividend can be altered directly (with the success or failure of this, to be assessed via the effects on inflation). It’s hard to isolate this effect and assess the success of the policy so, although this is the most direct way of pushing aggregate demand up or down, it is the last resort.

The net effect of the combination of a citizen’s dividend and a variable land value tax is similar to a national government changing the size of its deficit, which we would normally expect to be offset via interest rate changes, instead here the change in fiscal balance will reduce the required swings in interest rates. (This is particularly helpful at the zero lower bound, with negative rates or at very high interest rates where interest rates seem to be less effective and can negatively impact the economy or national debt burden.) It is intended to replicate and/or complement the functions some have argued should be taken up by fiscal councils. (Various papers, see below)

The proposal for a citizens dividend is included here because it is a simple way for central banks to inject money directly into the economy and ensure a basic minimum level of aggregate demand. The need for fiscal-monetary cooperation, and therefore ‘helicopter money’, has been highlighted by the recent crisis and it should always be able to increase demand as required. (McCulley, 2013 and Buiter, 2014) Affordability is obviously no issue for a currency-issuing bank but inflation is. The concept is that with the dividend rising slightly above the inflation target in normal circumstances then the other tools should be sufficient to curb inflation and if not the rate of increase can be lowered. Even if the dividend starts below the level of a full UBI it should eventually rise to the point where it becomes one.

This, therefore, interacts with the Land Value Tax (LVT) to create a surplus or deficit that behaves in the same way as a government surplus or deficit and so can stimulate or slow down the economy as required.

The other benefit of the citizen’s dividend/LVT combination is that a worry with a universal dividend payment would be that it simply gets swallowed up by landlords via rent. With this system the LVT is adjusted to ensure a specific and regular growth in land values. This means that if land becomes more profitable (because the dividend is being consumed by rent, for example) then capital will shift from other activities to land ownership, pushing up land prices. That forces the central bank to raise the LVT, stopping the landlords from capturing the rent. Land is forced to simply follow the normal rate of profit for any capital asset in the financial system. This also removes the unearned rentier income that landlords currently enjoy, leaving only unearned rents from financial assets generally as a problem. This is discussed later in this proposal.

Land Value taxes have been a topic of economic thought and discussion for over two centuries, from Adam Smith (Smith, 1776) to John Stuart Mill (Mill, 1848) to Henry George (George 1879), and they are appropriate here for several reasons. They can be collected under all circumstances as land cannot be hidden or moved offshore. This means it is always possible to remove excess currency from the economy via this method. They are also, by definition, regional meaning that the effective surplus or deficit will be different in different regions in a way that is directly proportional to what support those regions need. This is because the dividend is equal for each person but land values vary widely by region. This addresses a major flaw of “one size fits all” interest rate policy. This can help address the problems raised by New Economic Geography. (Krugman, 1991 and Krugman, 1998)

Land value taxes are also clearly separate from prices, in contrast to VAT for example, so raising LVT does reduce the available money in the system without adding to prices directly. It is also simpler to administer than something like income or wealth taxes as the taxable asset is immovable and easily externally assessed. Pairing the LVT with a citizen’s dividend also overcomes a common objection to land value taxation- that LVT isn’t based on ability to pay. The majority of citizens will own less than the average land value of land but will receive an equal dividend to everyone else and so the policy will be largely progressive. It may be that the policy ultimately results in more equal land ownership but as the combination of the dividend plus the LVT combines to achieve positive inflation (and hopefully a growing economy) in normal circumstances the total dividend payments given out will be greater than the total LVT collected in and therefore it should be extremely rare that someone who can’t afford to pay is deeply hurt by the system. Indeed under most circumstances, and certainly under current conditions, it would be deeply progressive.

Current interest rate changes tend to be most effective in increasing or curbing inflation via borrowing for purchasing land (commercial and residential) because these are long-term purchases where interest rates are a strong determinant of overall cost, as compared to short-term business purchases where the cost of borrowing is a less important factor. This is even more true with variable rate tracker mortgages where the current interest rate can quite directly affect the homeowner's finances. In this respect altering land value tax, as described here, is somewhat like putting every single piece of land on a tracker mortgage. LVT can, therefore, be an effective tool for slowing an overheated economy or providing support to a struggling one, as long as the Universal Dividend is providing a floor to aggregate demand and land values are following the business cycle.

This is the next challenge and is the area where the LVT interacts with the central bank annuities and interest rates. Interest rates can make land more or less expensive as many people borrow to buy housing or office space, affecting land values, but LVT should also be able to achieve the same effect fairly easily as it is also a regular payment that those purchasing land will have to consider before they buy. During a boom rental values for land should increase, as should appetite for taking out loans to purchase housing, this should raise prices on land. Raising LVT such that land values tend to rise below inflation will dis-incentivise land banking and also curb the rise in the cost of land during a boom. Conversely during a downturn lowering LVT can help avoid land value deflation and protect against people who have taken out mortgages going into negative equity. This means that LVT is able to eventually slow inflation, just as interest rates can, and therefore moderate the business cycle. It should also be able to filter through into the real economy as quickly as interest rate changes. A change in the next LVT payment can be set up to be as quick as a change in interest rates affecting the next payment on a tracker mortgage and much quicker than slower interest rate effects that are less direct. The concern, however, is interest rates and LVT working against each other- i.e. during a boom rental values may rise but if interest rates, and therefore return on capital, are rising generally as a percentage of capital values then, as yield is inversely proportional to price, land values would be unlikely to rise as much as if interest rates were constant. The system should, therefore, aim to achieve more stable and consistent interest rates.

The annuities address this long term consistency goal by, in effect, being very long-term interest rate targeting. They simply imply that over an infinite time horizon capital will be available for any profitable project even if capital is limited in the short run. They imply long run rentier income on capital of zero, although this is somewhat more nuanced than it first appears. If the price of the annuities is successfully kept constant, such that the yield is equal to the inflation target and inflation is, on average, at the target rate then the real interest rate on the (virtually risk free) annuity is zero but this doesn’t preclude risk-adjusted market rates rising such that investors are demanding an interest rate that, on average, gives a positive net return for risky assets. This risk premium would be charged at the discretion of the market.

Why target a long run risk-free real return of (or close to) zero? Partly because the current financial system is capable of creating credit as required for any project that creates a profit and there is no longer a limited money stock available to draw from. Even before this was the case economists had long talked about how as an economy advances the real return on capital should drop, this point was raised in Wealth of Nations (Smith, 1776) and in Keynes’ General Theory (Keynes, 1936). During the ‘great moderation’ period rates did slowly drop and have now been zero in many nations or in some jurisdictions there have been negative real interest rates. Given that this is the case and given that the fiscal tools of a dividend and LVT should be able to tackle over or under inflation there does not seem to be a need to grant a long-run real return on existing capital stock. As Keynes says in General Theory: we must aim for “the euthanasia of the rentier”. (Keynes, 1936) Indeed if investors do demand a risk premium across the market, such that most can be confident of a positive return, then, by definition, investment is not quite so risky, meaning that any risk premium charged will undermine itself, pushing the market back towards no rentier income. Moderating investor risk appetite and liquidity preference is a vital role played by the annuities.

Other yields will be valued in relation to the safety of perfectly liquid cash and highly safe and highly liquid annuities. Land will clearly not be as safe as these annuities but the fact that LVT will alter to ensure a positive nominal increase in land values will make them a relatively safe asset in comparison to market assets that don’t have this property. As such the nominal yield on land will not be the same as the central bank annuities but it will be related to it. A general increase in economic activity will push up rents on land and, because there is a relationship ensured between land values and rents via yield arbitrage with the central bank issued annuities, then as land becomes more attractive land values will also rise with those rents and therefore with the business cycle. The reverse is true during a recession. This ensures that the LVT tool is effective at curbing inflation by reinforcing the relationship between land values and inflation.

Put another way the annuities are anchoring long-term rates of return so that rising rents always mean rising land values, which ensures land value taxes can always tackle inflation and therefore moderate the business cycle. I.e. as rental values and risk appetite both rise in a boom, and as reserve requirements for annuities are raised, private investors will swop annuities for cash from central banks and the rising rents will ensure some level of this is invested in land, pushing up land values, causing an increase in LVT.

The use of annuities does not remove traditional short-run interest rate targeting from the Bank of England’s policy toolset but it is intended to mute it and reduce the severity and duration of swings in the short run rate. In normal circumstances it may be preferable to keep short term rates similar to the yield on the annuities and let the reserve requirements and LVT slow the economy as inflation rises. This is in line with Keynes’ comments on interest in General Theory about using redistribution rather than interest rates to moderate a boom. (Keynes, 1936 ch. 22)

It’s helpful at this point to talk about where these annuities fit on a yield curve. Are they at the long end because they have no maturity date and are, therefore, in effect, infinitely long-dated? I don’t believe this is the case. Instead, they are variable based on investor beliefs about when the central bank will succeed in hitting its price target for them. During a recession when liquidity preference is high this should be relatively easy and so the annuities are, in effect, very short-dated. In a boom when investors want to move away from safety and into risk assets it may require a considerable increase in the reserve requirements for annuities to achieve this target and so the date by which investors expect the central bank will achieve this target moves into the future and so the annuities move out along the yield curve. This would be a signal that the market is likely overheated. This doesn’t necessarily imply the bank should raise short term rates, if inflation is slightly above target but stable then it seems more sensible to let the LVT and the reserve requirements work. However, if inflation is considerably above target and rising then raising short term rates continues to be a tool available to central bankers.

Competition between central bank annuities, land and financial assets will allow risk appetite to be satisfied and rewarded during the boom with the level of risk appetite clearly shown in the difference between the yield on the safe assets vs the market assets that have no safety mechanism during a recession.

Therefore the reserve requirements are important for shifting risk between the core credit creating “high street” banks and private investors. At any reserve requirement below 100%, the core banks are technically capable of issuing an infinite amount of credit, just as they are now, however with each increase in reserve requirements the amount of that credit which is offset by safe (but 0% real yield) assets increases. Private investors will not have this burden. As such during a boom private investors will sell annuities to the banks, which will need them to meet reserve requirements, credit creating banks will then be in competition with financially constrained private investors for providing loans. Any risks taken on by the banks at this point will be increasingly offset by safe annuities while private investments will not. During a recession the process should invert- the banks will have been cushioned by their annuities and as private investors revert to wanting safety the price of annuities will rise, causing reserve requirements to drop. Banks will then sell annuities and buy up cut-price financial assets with their newly available credit, credit that is available because of the lower reserve requirements. This lines up with Roger Farmer’s work on stabling asset prices to protect the economy from swings in asset values. (Farmer, 2018)

The forces holding back the proliferation of loans will be similar in some ways to the current arrangement but will have some important differences. The overnight rate on reserves will continue to be a factor in restraining credit creation but this would ideally be more muted than at present and only used to tame very large economic booms or inflation spirals. Although the use of the overnight rate should, of course, be used as and when required, it is preferable to allow the other tools to kick in first. This avoids interest rates interfering with the LVT as much as possible but also creates more stable and predictable credit and capital yield conditions that can serve as a baseline for markets. This works because banks will still be operating on the same “borrow short-lend long” business model. A flat or negatively sloping yield curve is clearly a challenge for banks while an upward sloping one is more profitable. This gives us the doctrine of how overnight rates should interact with annuity reserve requirements. As reserve requirements are raised the banks must either swop reserves for annuities with the central bank or else swop deposits with private investors for annuities they hold. If the central bank does not intervene this should allow the overnight rate on reserves to rise above the yield on annuities. I.e. borrowing short and lending long is now more expensive for banks as they must issue both the deposits required to buy whatever long term asset they want plus the deposits required to purchase the relevant amount of annuities, all while the average return is lowered from the rate of return paid by the asset they are purchasing to the average of that asset and the annuities required due to reserve requirements. This happens at the same time as private investors have more liquid cash to invest due to the banks being forced to have swopped deposits for annuities from those investors. That causes the investors and the banks to be in competition and the private investors to be able to access a higher return from purchasing the same asset as they are not required to offset some of that purchase via annuities as the banks do via the reserve requirement. This pushes risks and returns from the core banks to private investors during a boom when risk appetite is high.

In addition during a boom LVT and rent changes will be reducing the profitability of businesses and making home purchases more expensive and less attractive, slowing the uptake of new loans. At the same time the selling of annuities by private sector investors to core banks will ensure liquidity amongst financially constrained investors who can then compete with credit creating banks for providing loans and investment. Much of this money may get trapped in the financial system due to rising asset prices and increasing velocity of financial transactions, as we saw with QE, however, it should also create an alternative source of funding for borrowers that follows slightly different dynamics to the core banking system.

If inflation is overshooting the yield on annuities/the target interest rate (and possibly the overnight interbank rate is rising) the rising reserve requirements impose a real short term cost on the credit creating core banks due to buying annuities that are yielding less than the overnight rate and this must be offset if banks wish to maintain the same profits as private investors.

How and when banks choose to create additional credit, above and beyond the total amount they had previously lent out, will depend on expectations, inflation and reserve requirements- if banks predict a long boom ahead before any recession then the economically rational choice would be to pick up the riskier (but higher interest rate) assets. The higher the reserve requirement, and the higher the overnight rate is, the higher the interest rate banks would want to charge for this additional credit. That discourages new credit creating borrowing. Conversely if banks expect a recession to be imminent then a better strategy is to sit on the annuities they have and wait for a recession to push up liquidity preference, forcing up annuity prices.

Put together this implies that the market rate of interest may well still rise during an economic boom but the rise will, hopefully, be shifted towards riskier assets while safer assets remain closer to the target long run rate as indicated by the yield on annuities. The aim would be to aim for a situation where, as the boom progresses, the market rate of interest on riskier assets would increasingly diverge from safer assets rather than all interest rates being forced up via a change in the overnight rate due to central bank action. This perhaps relies on economic actors being more rational than they really are and the system may not work as neatly described here. Even if that is the case it, at worst, behaves as well as our current system with the interest rate rising for everybody but, crucially, with the risks to core banks and private investors being asymmetric with core banks far more protected due to the annuity reserve requirements. The costs of the land value tax would also be asymmetric, concentrating on booming geographical locations and affecting those that aren’t enjoying the boom far less.

A reasonable concern at this point would be whether during a recession interest rates can be dropped sufficiently to allow a return to growth? After all how far can overnight rates on reserves drop below the coupon on the annuities before the reserves are simply used to buy more annuities instead? Also if there’s an infinite supply of safe assets available won’t investors simply all buy those instead of market assets?

This is offset by three factors- firstly the market rate of interest can drop from the height of the boom into the recession. It may be that outside of very high levels of inflation the central bank caps the overnight rate at slightly above the annuity yield/inflation target but it should be allowed to rise above it and so will be able to fall below it during a recession. This reshapes the yield curve and allows banks to create reserves more freely.

Secondly, the fiscal tools now available to the central bank mean that more cash can be placed directly into people’s pockets. If the LVT fails to achieve this on its own, hopefully this would be unusual but it could happen in a very large crash like 1929 or 2008 (it may happen if the extra cash is simply used to buy annuities rather than being spent in the economy for example), then the universal dividend can be raised to whatever level is necessary to achieve the target level of inflation. If the economy overshoots the target then the LVT can pull that same cash back out of the economy as required.

Thirdly core banks will be effectively ‘bailed in’ during a recession as they can swop overpriced central bank annuities (now in high demand as private investors search for safety and that the core banks have a ready supply of as reserve requirements are lowered) for cheap deals on market assets or simply money from investors, destroying credit money and reducing the core banks liabilities. Core banks would, in effect, be “buying the dip”. They can as afford to take over loans to companies at their fair value so as to avoid a credit crunch scenario for companies that are otherwise solvent and profitable. The central bank can continue to keep the overnight rate as low as is possible by purchasing annuities itself. Between these factors and the low reserve requirements during a crash, it seems reasonable that enough of the banking system would function in order to ensure credit and investment is still freely available as required.

These three tools should interact such that there is always sufficient aggregate demand to hit the inflation target but that any increase in inflation will cause annuity prices to fall and land rents to rise which causes reserve requirements to rise and, as land values rise, land value taxes to increase. Overall inflation will, therefore, still be curbed. Interest rates for borrowers will likely still rise but this represents an increase in risk appetite and not a rise in the risk-free yield on assets. Risk-adjusted yields will not have actually risen, it is simply that investors will be at a stage where they are willing to take that risk, as you would expect during a boom.

The price investors pay for higher yields during the boom (and banks also pay for any purchases they make with their credit creation capacity other than annuities, or to some extent land/mortgages) is that they are left in the lurch during the next recession. This interplay of risk, land value taxes and a consistent base for aggregate demand should allow for inflation to be kept close to the target with only small adjustments in short-term interest rates and without the risks of hitting the zero lower bound or hyperinflation.

The proposal does this while addressing regional and individual inequality and asset price volatility. This is achieved with low interest rates, which in turn facilitate government and infrastructure investment.

There are many other topics to consider in relation to this proposal that there isn’t space to discuss here. Some of the principal ones I would like to explore are:

-How would this interact with banking regulation and the cycle of high regulation after a recession and gradually lower regulation during times of moderation? How would core bank insolvency and bankruptcy be dealt with to reduce or remove both systemic risks and moral hazard?

-How does this system function in stagflation/hyperinflation and other atypical scenarios?

-How does this interact with new economic geography and agglomeration effects?

-What are the international implications?

-How does this system interact with government fiscal policy, from austerity to fiscal rules to deficit bias? Also with national debt.

-And should the inflation target be changed to better accommodate this system?

References

‘Helicopter Money: Or How I Stopped Worrying and Learned to Love Fiscal-Monetary Cooperation’ Paul McCalley, 2013

‘The Simple Analytics of Helicopter Money: Why It Works — Always’ Willem H. Buiter, 2014. Economics Discussion Papers, No 2014–24, Kiel Institute for the World Economy.

‘An Inquiry into Nature and Causes of the Wealth of Nations’ Adam Smith, 1776

‘Principles of Political Economy with some of their Applications to Social Philosophy’ John Stuart Mill, 1848. Particularly Book V, Chapter II: On the General Principles of Tax

‘Poverty and Progress’ Henry George, 1876

‘Increasing Returns and Economic Geography’ Paul Krugman, 1991. Journal of Political Economy volume 99. №3.

‘What’s New in New Economic Geography’ Paul Krugman, 1998. Oxford Review of Economic Policy Volume 14. №2.

‘A General Theory of Employment, Interest and Money’ John Maynard Keynes, 1936. In particular chapters 22 and 24.

‘Pricing Assets in a Perpetual Youth Model’ Roger A. E. Farmer, 2018. Review of economic dynamics 30 p106–204

Also relevant:

‘Aggregate fiscal policy under the Labour Government 1997–2010’ Simon Wren-Lewis 2013. Oxford Review of Economic Policy Volume 29, Issue 1, p25–46

Various relevant papers have discussed fiscal councils:

Calmfors, Lars (2003), Fiscal Policy to Stabilise the Domestic Economy in the EMU: What Can We Learn from Monetary Policy?, CESifo Economic Studies, 49,319–353.

Debrun, Xavier (2011). “Democratic Accountability, Deficit Bias, and Independent Fiscal Agencies,” IMF Working Papers11/173, International Monetary Fund.

Von Hagen, J. (2010). ‘The Scope and Limits of Fiscal Councils’, Paper to Conference on Independent Fiscal Institutions, March 18–19, Fiscal Council Republic of Hungary, Budapest.

Leith, C and Wren-Lewis, S (2006), Fiscal Stabilisation Policy and Fiscal Institutions, Oxford Review of Economic Policy, 21, pp 584–597

Wren-Lewis, S. (2003), The compatibility between monetary and fiscal policies in EMU: a perspective from the Fiscal Theory of the Price Level, in Monetary And Fiscal Policies In EMU: Interactions And Coordination, ed(s) Buti, Marco, Cambridge University Press.

Wren-Lewis, S (2011), Comparing the delegation of monetary and fiscal policy, Oxford Economics Department Discussion Paper №540

Calmfors, L and Wren-Lewis, S (2011), What should fiscal councils do? Economic Policy Vol. 26, pp 649–695 and Oxford Economics Department Discussion Paper №537

Debrun, Xavier, David Hauner, and Manmohan Kumar (2009), “Independent Fiscal Agencies”, Journal of Economic Surveys 23(1): 44–81, February.

Hallerberg, M., Strauch, R. and J. von Hagen (2009). ‘The Design of Fiscal Rules and Forms of Governance in European Union Countries’ in Ayusi-i-Casals, J., Deroose, S., Flores, E. and L. Moulin (eds), Policy Instruments for Sound Fiscal Policies, Palgrave Macmillan, London.

Wren-Lewis, S (2011b) ‘Fiscal Councils: The UK Office for Budget Responsibility’, CESinfo DICE Report 3/2011