Missing The Risk

Missing The Risk Paper


Section 1: Introduction.

The Office of Budget Responsibility, (OBR), was created by George Osborne in 2010, with its main duty being to examine and report on the sustainability of the public finances. Central to the OBR’s remit is the requirement for it to produce independent forecasts of the economy and public finances, setting out the key assumptions made, and giving an analysis of the risks surrounding the economic outlook.

The purpose of the present study is to examine the handling of risk in the OBR’s approach to forecasting the economy: and in particular, to assess whether the OBR is adequately assessing the full range of risks surrounding its economic forecast.

One major conclusion of this study is that the OBR’s treatment of risk is grossly inadequate, and likely to give an unduly reassuring impression. OBR’s main risk assessment is based upon historic forecasting error: i.e., the range of error in past forecasts of quantities like GDP made by various forecasting agencies. But this type of forecasting error is only one of the potential risks. Factors which are largely neglected in the OBR reports include:

a) the likelihood of future shocks to the economy.

b) the likely robustness of the economy to withstand a possible future shock.

c) feedback effects, because OBR bases some of its important assumptions on the view taken by markets on factors like future interest rates, and in turn the markets will themselves be influenced by OBR’s forecasts.

d) the fact that, since we are now in many ways in new economic territory, (for example, as regards the implications of the policy of quantitative easing), the past is not necessarily a useful guide to the future.

But this is not just a question of “could do better”. There are fundamental problems about producing forecasts in the kind of environment in which the OBR must operate – an environment where the outcome of the quantity being forecast is itself to a large extent under policy control. This is because, as will be argued later in this paper, a rational central forecast in such an environment will almost always automatically assume the success of policy. In such an environment, therefore, the central forecast itself is of relatively little interest. What is of much more value is an analysis of the risks surrounding the likelihood of the central forecast being achieved.

So the second main conclusion of this report is that the OBR should be radically changing its emphasis. It should be putting much less effort into forecasting, and much more into assessing the potential vulnerabilities to which the economy is subject, and the potential risk factors which could prevent government achieving its economic policy aims.Such a change in emphasis might be less superficially reassuring than the OBR’s current reports: but it would be much more valuable. It would be valuable for the public, because they would feel that they were getting a properly detailed examination of risk. But it would be of particular value to government, because it would almost inevitably focus attention on vulnerabilities in the current economic structure which should be rectified, but are currently in danger of being ignored.

The structure of this paper is as follows:-

• Section 2 gives a brief outline of OBR’s economic forecasting methodology, bringing out the sources of the key assumptions.

• Section 3 summarises, and assesses, the way in which OBR currently handles the question of risk.

• Section 4 looks at the problem of forecasting in a heavily policy influenced environment, and considers the implications.

• Section 5 makes specific suggestions about how OBR should change its approach.

Section 2: The OBR Approach to Forecasting.

The OBR’s remit and status is set out in the government’s Charter for Budget Responsibility, (Treasury, 2011). The primary duty of the OBR is to examine and report on the sustainability of the public finances, (Charter, para 4.2), and the OBR is, by statute, independent. It does, however, have some particular duties laid on it by the Charter. It has to produce regular forecasts of the economy and public finances. And it has, specifically, to address the question of whether the government is on target to meet its fiscal mandate: that is, it has to assess whether the government is meeting its objective of balancing the current budget, on a rolling five year basis adjusted for the economic cycle. Further, in each report published under its main duty, it has to explain the factors taken into account when preparing the report, including the main assumptions and risks: (Charter, 4.7).

This paper is particularly concerned with the way in which the OBR produces its forecast of the economy, and how it assesses the associated risks – rather than with the OBR’s forecast of the public finances. The OBR’s approach to economic forecasting is described in detail in its Briefing Paper Number 3, (OBR, 2011). As the OBR notes, its approach to economic forecasting is somewhat different from that adopted by other independent forecasters. Because of the detail required as input to their forecast of the public finances, the OBR forecast contains more economic variables than most other forecasts: and because of the five year period implicit in the government’s fiscal mandate target, the OBR forecast covers a longer period than most others.

The OBR has adopted the detailed Treasury macroeconomic model as one of its main tools in producing its economic forecast. The OBR maintains a version of the model independently of the Treasury, and has made it clear that it is prepared to modify the Treasury model if it feels the Treasury version no longer meets its requirements. So, in principle, the OBR’s use of the Treasury model is still consistent with its independent role.

It is very important, however, to appreciate that the key determinant of the OBR forecast is not the econometric model itself: but rather the set of judgements and assumptions which the OBR makes, and which feed into the model. This is a point which the OBR themselves make very clearly, (see, for example, Briefing Paper 3, para 2.2 and 2.16ff), where the judgemental nature

of the OBR forecast is stressed. The advantage of the econometric model is that it enables the implications of a given set of judgements and assumptions to be worked out in the detail required for making fiscal forecasts: and also, crucially, because it means that the OBR forecast is constrained to paint an internally consistent picture. (That is, the model should prevent the OBR from making assumptions which are internally inconsistent – at least in terms of the internal logic of the model.) These are important advantages of using the econometric model. But the key determinants of the OBR forecast are the judgements and assumptions which the OBR make. We now consider the set of judgements which the OBR is making.

Probably the two most important judgements which the OBR make are those concerning the so- called output gap, and the trend growth rate of potential output. Some explanation is required of these concepts: for more detail, see for example, OBR’s Forecast Evaluation Report, (OBR 2012b), para 2.80. The potential output of the economy is defined as that level of economic activity, (that is, GDP), which is consistent with maintaining stable inflation. The output gap, (which is usually expressed as a percent of GDP), is the difference between the current level of activity in the economy, and the level of potential output. If the output gap is negative, then this means that there is spare capacity in the economy: i.e., economic activity could be increased without generating inflationary pressure. Conversely, if the output gap is positive, the economy is overheating.

As well as these key judgements on the size of the output gap, and on the trend growth rate of potential output, the OBR make a further vital assumption: namely, that the Bank of England, (BoE), manages monetary policy in such a way that the output gap closes over the forecast period. In other words, the OBR assume that the BoE manages interest rates so that, round about the end of the forecast period, inflation is stable at the government target rate of 2% CPI inflation per annum: and that the economy is then operating on the trend line of GDP growth: (see, for example, Briefing Paper 3, para 3.30.)

These two key judgements, (on the output gap, and the trend of potential output), and the assumption about the success of monetary policy, effectively determine the basic outline of the OBR economic forecast, in terms of where GDP and inflation will be around five years ahead. Of course, there is a lot more to the OBR forecast than this: for example, in forecasting what the path of GDP will be between now and the end of the forecast period: and in making assumptions about, or deriving from the model, the whole set of detailed information which OBR require. But the essentials of the OBR forecast are determined by the two key judgements and the monetary policy assumption.

A crucially important point to note is that one of these three elements is the assumption of the success of monetary policy. In other words, success of a key aspect of government policy is taken for granted as a central part of OBR’s forecasting methodology. This is a point on which there will be much more to be said in section 4 of this paper. It might at first sight appear irrational for an independent forecaster to be assuming the success of an important aspect of government policy, when the forecasts being produced are intended to be an independent validation of the credibility of that very policy. It will be argued in Section 4, however, that the OBR’s monetary policy assumption may not be as irrational as it might first appear, given the very real conceptual problems attached to forecasting in a heavily policy influenced environment. But, as will be argued, these problems have profound implications for the status which should be attached to the OBR’s projections, and also for the kind of work OBR should be doing.

Other assumptions made by OBR in the forecasting process are as follows: (this is not a complete list):

• Oil prices: that these move in line with the current values of Brent Crude futures prices.

• World interest rates: as predicted by current values of relevant interest rate futures. • UK gilt edged interest rates: as predicted by current values of relevant interest rate

futures. • Equity prices: grow in line with nominal GDP. • House prices: as projected by the median of independent forecasters.

• Exchange rates: determined by model, in line with what is known as the “uncovered interest rate parity condition”: (that is, exchange rates are assumed to move to eliminate the effect of interest rate differentials between the UK and the rest of the world.)

• Unwinding Quantitative Easing: OBR assume that Quantitative Easing will begin to be unwound, (that is, that the BoE will begin to sell off the UK Gilt edged securities it has purchased through quantitative easing), as soon as Bank Rate rises above 1%, with sales evenly paced at £10 billion per annum thereafter.

[Sources: mostly from OBR Briefing Paper 3, para 3.3, apart from the assumption on Quantitative Easing, which is from OBR 2013, para 4.23.]

It is worth making two points about this list of assumptions.

a) For certain important variables, (e.g., UK and world interest rates, and hence also exchange rates), the OBR is relying upon the view which the markets are currently taking about what will happen in the future. This raises the potential for feedback effects between OBR forecasts and market views, feedback effects which could be potentially destabilising. This will be discussed at greater length in section 4.

b) In at least one very important respect, (the assumption about unwinding quantitative easing), the OBR are again assuming the success of government policy. In other words, it is being assumed that quantitative easing will be unwound, (so the government is not actually monetising a large part of its debt), and that this can be done without disrupting the future path of interest rates as currently predicted by the market.

This concludes the description of the relevant parts of the OBR’s approach to forecasting. This paper is primarily concerned with the OBR’s approach to risk, rather than with a general critique of OBR’s central forecast. But before leaving this section, it is worth making one comment about OBR’s approach to forecasting. Namely, that it is surprising that OBR give such a central role to the concept of potential output. What this means is that OBR are basing their forecast primarily on a judgement about a quantity which cannot even be historically observed. This makes it extremely difficult, if not impossible, to do a proper post-mortem on the OBR forecast, if forecast GDP growth does not, in the event, correspond with what actually happens. Of course, it is very convenient for politicians to express their fiscal mandates in terms of concepts which cannot be unambiguously measured: (witness Gordon Brown’s redefinition of the economic cycle as it became clear he might breach his Golden Rule.) The OBR certainly cannot escape consideration of the output gap, given the duty laid upon them to assess whether the government is on course to meet its fiscal mandate. But it is doubtful if they should have taken on board the concept of potential output to such an extent that they actually placed it at the centre of their forecasting approach.

Section 3: The OBR’s Approach to Risk.

As already noted, one of the requirements laid upon OBR in reporting on the sustainability of the public finances is to explain the factors taken into account in preparing their report, including the main assumptions and risks. In the OBR’s March 2013 Fiscal Outlook Report, three main techniques are employed for assessing risk:

a) by looking at past forecast errors.

b) sensitivity analysis: that is, investigating how much OBR’s central forecast would change if some of the key judgements underpinning it were altered.

c) scenario analysis: that is, examining some alternative economic scenarios.

(Further details on these techniques can be found in OBR 2012a)

In line with the first of these approaches, several of the charts showing forecasts of important variables are presented as fan charts, where the central forecast is surrounded by an increasing fan showing the probabilities attaching to different ranges of outturns: the width of the fan reflects experience with past forecast errors.

In line with the scenario analysis approach, material presented in Chapter 5 of the March 2013 Fiscal Outlook looks at the effect on the government’s target fiscal aggregates of

a) various combinations of the assumed size of the current output gap, (5 variant options), with various combinations of the assumed date when the output gap finally closes, (again, 5 variant options).

b) 5 variant assumptions about the future interest rate at which the government borrows.

c) 4 different variant options in the assumed coefficients used in cyclically adjusting the fiscal aggregates.

Finally, in line with the scenario analysis approach, two broader illustrative scenarios are analysed in Chapter 5 of the March 2013 Fiscal Outlook: both scenarios are based on different reactions to a 15% depreciation in the Sterling exchange rate relative to the central forecast. As OBR note,

“We stress that these scenarios are not intended to capture all possible ways in which the economy might depart from the central forecast: and we do not attempt to attach particular probabilities to these occurrences.” (Fiscal Outlook, para 5.36)

There are a number of very significant weaknesses in the above approach. In particular:

a) As regards the fan charts based upon past forecast errors, it is important to remember that forecasting error, (i.e., the kind of random error one would normally expect in forecasting any system operating under fairly stable control), is only one of the potential sources of error. Further, it is by no means obvious that the past is a reliable guide to the statistical distribution of forecasting errors in the present or near future. This is particularly the case since economic circumstances over most of the period when past forecast errors would have been assessed are very different from what they are now. With interest rates at unprecedentedly low levels, it is uncertain whether the relationships on which the OBR economic model is founded will continue to hold over the forecast period.

b) The sensitivity analysis which the OBR report on is very limited. It is limited not only because of the very small range of variant possibilities which the OBR consider: it is also limited in the nature of the analysis conducted. As the OBR note, they have carried out the sensitivity analysis on a ready-reckoner basis: that is, this effectively amounts to saying – what is the effect if one assumption is varied, while everything else is held constant? Real life, however, is likely to be much more complex, with dislocations of the central forecast likely to involve departures of a number of variables from their originally assumed values. So the limited form of sensitivity analysis conducted by the OBR is likely to significantly understate the actual sensitivity of the OBR forecast to variations in assumptions.

c) The scenario analysis undertaken by OBR is meant to represent a more integrated approach to evaluating the effect of different economic scenarios. But the number of scenarios considered, (only two, both relating to exchange rate movements), is extremely limited.

The analysis of forecasting error, and the sensitivity and scenario analysis conducted by OBR, thus give no real indication as to the potential uncertainty surrounding the central forecast. In particular, the following are effects and contingencies which would need to be considered in any proper assessment of risk and uncertainty.

Potential feedback effects. As noted in the preceding section, the OBR look to the futures market as the source for certain key assumptions: in particular, interest rates, but also crude oil prices. But here we encounter a potentially destabilising paradox. The forecasts produced by OBR must in themselves be an important influence on the market’s view of the future: after all, OBR is intended to be an authoritative and independent forecaster, whose views undoubtedly carry considerable weight. So when the market is forming its view on future interest rates, it will in part be influenced by OBR, which in turn takes its assumption from the market…. . It is clearly very possible in these circumstances that the market and OBR might become caught up in a cycle of misleading reassurance: and that, equally, feedback effects could lead to considerable instability if such a cycle broke down.

Quantitative Easing risk. As noted in the previous section, OBR assume that Quantitative Easing, (QE), will start to be unwound as soon as Bank Rate reaches 1%, (which, in OBR’s March 2013 forecast, is projected to happen in 2016, based upon current interest rate futures.) There must, however, be huge uncertainty attaching to the various ramifications of this assumption. We are, after all, talking about the process of unwinding what Andy Haldane, the BoE’s Director of Financial Stability, described as recently as June 2013 as “the biggest bond bubble in history.” What will happen in relation to QE unwinding must be inherently uncertain – for one thing, this is completely new territory. For example:

If the market ever came to the view that the government was not going to unwind QE, then this would amount to the view that the government was monetising its debt – such a view would have profound and unprecedented consequences on inflation, and confidence in Sterling.

On the other hand, once QE unwinding actually starts, then this will remove the support to current bond prices supplied by the anticipation of future QE– leading to the possibility of a collapse in bond prices, (i.e., increase in interest rates), which might far exceed what would normally be expected from the relatively modest scale of gilt sales, (£10 billion per annum), which OBR forecasts.

Overall, the risks surrounding QE, (not just in the UK, but worldwide), are so considerable that the BoE’s Andy Haldane singled this out, in evidence to the Treasury Select Committee, as the biggest risk to global financial stability.

Effect of external shocks. There is some limited discussion in the OBR documents of the potential effects of future external shocks, (e.g., a messy outcome in the Euro zone), but there is nothing like a full discussion. And OBR have a tendency, when discussing such possibilities, to incorporate the effect in terms of an adjustment to their central forecast, but not in terms of uncertainty. The following quotation illustrates this:-

“As set out above, we expect constrained real income growth, ongoing dislocation in financial markets, the fiscal consolidation and weak global growth to limit the rate of growth over the medium term. As a consequence, we expect the output gap to narrow at a relatively gradual rate…” (OBR 2013, para 3.34).

It is perfectly correct for OBR to adjust their central forecast in the light of their view of factors like dislocation in financial markets, and uncertainties in global growth. But uncertainty in such factors will affect not just the mean of the forecast distribution: it will also affect the variance, (that is, the spread of outcomes), around that mean. This increased uncertainty should be reflected in any realistic assessment of risk: OBR fails to do so.

Failure to assess systemic risk. As the events of 2008 illustrate, systemic risk, (i.e., the risk of the collapse of a large part of the financial system, with corresponding catastrophic impact on the economy), is not just a theoretical possibility. And yet there appears to be no assessment of the factors surrounding systemic risk in the reports produced by OBR. Such an assessment should at least include consideration of the following:

a) an assessment of the likelihood of a trigger, (like a messy resolution of the Euro zone problems, or the collapse in an asset bubble), which might conceivably start a chain of events potentially leading to systemic collapse.

b) an assessment of how robust the UK economy and public finances are to withstand a major shock.

Any realistic analysis of the second of these factors would presumably conclude that, with public sector net debt now at 76% of GDP and rising, compared with 36% immediately before the start of the 2008 crisis, the UK public finances are not well placed to withstand a further major shock. In other words, any realistic assessment would conclude there is a significant systemic risk to the UK economy.

Overall, in the light of the above, the conclusion appears inescapable: namely, that the OBR does not come close to fulfilling the requirement in its remit to provide an adequate assessment of the risks and uncertainties surrounding its forecast.

Section 4: The Status of Forecasts in a Policy Influenced Environment.

The OBR’s concentration on producing its central forecast, and the limited attention paid to assessing the risks and uncertainties surrounding that forecast, are doubly unfortunate because,

as will be argued in this section, central forecasts mean relatively little in relation to a heavily policy influenced environment like the economy.

Suppose that we are trying to forecast the outcome of a particular process, (like the economy), where there is an agent who has a specific policy objective for the outcome of the process, and where that agent has powerful levers that can be used to influence the outcome: (like the BoE, with its target for inflation, and its monetary policy control lever.) Then for an independent forecaster, the most rational choice for the central forecast will often be that the desired policy objective is achieved. This is because, if it is obvious to the independent forecaster that the process is currently heading for, say, an undershoot – then this will be equally obvious to the controlling agent. So the forecaster has to assume that the agent will take corrective action. Of course, this action may be too much, or too little: or other factors may come into play. But the important point is that there is normally no reason for an independent forecaster to assume that the outcome will either consistently overshoot or undershoot.

In other words, the normal rational decision made by an independent forecaster is likely to be that the required policy objective is attained. The only circumstance where the central forecast is likely to differ from this is where there is evidence the process is currently heading off course, and where the forecaster has firm evidence that the agent’s control lever has been rendered ineffective. This combination of circumstances is likely to be relatively rare.

This feature of forecasting in a policy influenced environment is entirely consistent with the fact that, as seen in Section 2, the OBR assumes that the BoE achieves its inflation target by the end of the forecast period: and with the assumption that quantitative easing will be successfully unwound. Making such assumptions is in line with OBR acting rationally: and does not in itself imply that OBR’s independence is compromised.

Since, in a policy influenced environment, the normal rational central forecast is that the policy objective will be achieved, this means that the central forecast in itself is of relatively little interest. But the corollary is that what is really important is the assessment of the risks surrounding the central forecast. Consider the following rather dramatic imaginary example. Suppose that the White Star Shipping Company, in order to reassure potential passengers on its new liner Titanic, had set up an independent White Star Forecasting Office, (WSFO), to validate the shipping company’s mandate that the new liner would be the fastest and most reliable on the North Atlantic. The rational central forecast produced by the WSFO will be that the Titanic will arrive in New York at around the scheduled time. This forecast in itself is of extremely limited value to the traveller, although the shipping company will make great play of it in its advertising. An additional analysis produced by the WSFO, showing the probabilities of the Titanic being earlier or later than the scheduled arrival time by differing amounts, and based on past experience of the punctuality of similar liners, is of some use to passengers to enable them to plan their ongoing connections: but in practical terms is little more than a convenience. What would really be of interest to the concerned traveller would be if the WSFO produced an authoritative assessment of iceberg conditions in the North Atlantic: and also produced a technical study on the chances of the ship surviving an impact. And such analyses would not merely be reassuring, (or otherwise), to passengers: the existence of an authoritative report highlighting the dangers would itself inform the shipping company about potential dangers it might have underestimated – and greatly reduce the risk of a catastrophic outcome.

Going back to the OBR, what is seen from this example is that the OBR has got its priorities precisely the wrong way round. It puts huge effort into elaborating a central forecast which is of little inherent value, because it depends on key assumptions which assume the success of policy outcomes. But on the other hand the OBR largely neglects the important role which it should be playing, of assessing the major risk factors surrounding the economy.

This situation is convenient for politicians – since the central forecast will almost always be reassuring, and the risks, not analysed, are therefore underplayed. But an independent body like OBR should not be making life easier for politicians, and perhaps unwittingly giving them a false sense of security. The OBR should change its approach: some of the changes it should make are considered in the next section.

Section 5: Suggested Changes.

It has been argued in this paper that the OBR put too much effort into producing its central economic forecast, and too little into assessing the risks around that forecast. In fact, in a 2011 briefing paper on forecasting the economy,(OBR 2011), the OBR itself raised the question of whether it should undertake macroeconomic forecasts at all: as it said then, (para 3.144):

“One might ask why the OBR should produce macroeconomic forecasts at all and why we should not use those of the Bank of England or other non-governmental bodies that produce forecasts for the UK economy.”

OBR, quite reasonably, concluded that it did need to produce a macroeconomic forecast, because no external forecast were available which covered the range of variables and timescale the OBR needed in order to assess the government’s fiscal targets. But the quotation is relevant because it seems to indicate that the OBR itself does not attach the same weight and status to its central forecast which some commentators, and politicians, are inclined to do.

In order to reduce the danger of undue weight being given to the central forecast, it would be advisable if OBR were to alter the terminology it uses to describe its central forecast: describing this more as a projection dependent on certain conditioning assumptions, rather than as a forecast. A suitable form of words might be:-

“This is our central assessment of the path of future economic development, based upon our judgement, but conditional upon certain key assumptions: namely

c) that monetary policy is successful in achieving the target rate of inflation by [date].

d) that quantitative easing starts to be unwound during the forecast period, and that this takes place in an orderly fashion, with no distortion in interest rates from the path currently implied by the futures market.

e) and that there are no profound external economic shocks.”

In placing greater emphasis on assessment of risk, it is suggested that there are a number of steps OBR should take. These are as follows.

• First of all, in presenting the results of its macroeconomic projection, the OBR should, as a standard approach, give long time series showing historic values of the key variables, and also the OBR projected values. It would also be useful if this information was presented graphically. This would make it easier for the user to assess whether the OBR projections implied any implausible looking changes in trend: and also whether the relevant variable was moving into an area which implied systemic risk. For example, time series showing the growth of different forms of credit, relative to

GDP, would have acted as flashing warning lights before 2008: (though, of course, OBR itself did not exist then.)

• Secondly, OBR should reconsider what variables it includes in its macroeconomic model: it may be useful, from the point of view of risk assessment, to include some variables which, from the point of view of driving other parts of the model, may not be strictly necessary. The following is one example to illustrate this point. The international assets and liabilities of the UK break down into a number of different categories – primarily relating to direct investment, portfolio investment, other investment, and financial derivatives. The assets in the first three of these categories generate dividend and interest payments which contribute to the current income flow on the current account of the balance of payments. However, the assets and liabilities in the fourth category, namely, financial derivatives, do not generate such income payments: (what counts as an asset or a liability for a financial derivative contract is the amount which the relevant party to the deal would have to pay, or be paid, to be relieved of the contract.) So in the OBR model, what is included as UK assets and liabilities are assets and liabilities excluding financial derivatives: and an assumption is made about the rate of return on these assets and liabilities, which the model uses to generate projected investment income.

While the exclusion of financial derivatives from the OBR model can be justified on model operational grounds, their omission means that something which is potentially very important from a risk point of view is excluded. The following chart illustrates this point.

from Jim Cuthbert paper

(source: historic data from ONS Pink and Blue Books: forecast data derived from OBR 2013.)

What the two upper lines in the chart show are total UK assets owned by foreign residents, and total foreign assets owned by UK residents, expressed as a percentage of UK GDP: (in the later years shown the first of these quantities has been slightly greater than the second.)

Also shown in the chart, in the two lower lines, are the above quantities excluding financial derivatives. These are the quantities which are modelled by OBR: and the chart shows, for the lower lines as from 2013, the values implied by the OBR March 2013 forecast.

If we think of the UK economy, in its international relations, as being rather like a very large bank, (a useful analogy made by Nickell, 2006), then this chart is one way of looking at the balance sheet of “Bank UK”. Clearly, the view we might take of the potential stability of Bank UK is rather different looking at the upper lines in the chart, as compared with the lower lines, which is what the OBR focuses on. For a fuller appreciation of the potential risks facing Bank UK, the OBR should be considering, and modelling, the upper as well as the lower lines in the chart.

(Note that there are some discontinuities in the ONS series for financial derivatives liabilities on which the above chart is based: but these do not invalidate the point being made.)

The above suggestions would imply useful improvements in OBR’s ability, and the ability of users of the OBR projections, to assess potential risks to the system. In a sense, however, these changes would still be relatively cosmetic, compared to the major effort OBR should be making: which is to put serious resources into identifying and analysing the major sources of instability and risk for the UK economy: and into assessing the likely robustness of the economy in the face of potential future shocks.

The OBR clearly commands considerable, and talented, resources. Redeploying these resources into a better assessment of risk would be immensely valuable. In particular, there are disturbing indications that policy makers have not learned the lessons of 2008: and that, despite the increased vulnerability of “Bank UK” in many respects since 2008, the primary current policy response is to get things “back to normal” by inflating further asset bubbles: (for example, through quantitative easing, or through recent government initiatives in the housing market.) As things stand, the OBR, by concentrating on producing a central forecast which to a major extent takes the success of government policy for granted, is actually proving a barrier to fundamental change. If it were to put much more effort into a proper assessment of risk, then the effect would be the opposite, and beneficial, one of acting as a stimulus for change in the UK economy.

Section 6. Conclusion

In this paper, it has been seen how

a) the OBR pays inadequate attention to the assessment of risks surrounding its central forecast of the economy.

b) OBR takes as a given in making its central forecast the success of certain key aspects of government policy.

The second point does not necessarily mean that OBR is failing to act independently – since it is a feature of attempting to forecast in a heavily policy influenced environment that a rational independent forecaster will often assume the success of policy in their central forecast. But the corollary is that the central forecast in such an environment in itself means relatively little.

Accordingly, the main recommendation of this paper is that OBR should radically shift the balance of their effort. On the one hand, OBR should put less effort into the production of their central forecast: and it should be clear, in publishing its results, that what it is producing is essentially a projection based on certain conditioning assumptions. On the other hand, it should put much more effort into its assessment of the risks surrounding the UK economy: in particular, in identifying the main potential shocks to the economy, and in assessing the economy’s likely robustness to withstand any shock.


H.M. Treasury, (2011): “Charter for Budget Responsibility”: April 2011.

Nickell, S., (2006): “The UK current account deficit and all that”: Bank of England Quarterly Bulletin, Summer 2006.

OBR, (2011): “Forecasting the Economy”: Briefing Paper No. 3: October 2011. OBR, (2012a): “How we Present Uncertainty”: Briefing Paper No.4: June 2012. OBR, (2012b): “Forecast Evaluation Report”: October 2012. OBR, (2013): “Economic and Fiscal Outlook”, March 2013

Join the Discussion

Your email address will not be published.

Help keep our journalism independent

We don’t take any advertising, we don’t hide behind a pay wall and we don’t keep harassing you for crowd-funding. We’re entirely dependent on our readers to support us.

Subscribe to regular bella in your inbox

Don’t miss a single article. Enter your email address on our subscribe page by clicking the button below. It is completely free and you can easily unsubscribe at any time.