Marc Robinson's Blog, page 3

Unit Cost Budgeting? (1/4)

Funding public services based on output unit costs is a great idea. However, it only works for selected government services. The idea that the entire government budget should be based on output unit costs is therefore misguided. This series of four blog pieces looks at why this is a tool with strictly limited application in government.

An example of unit cost budgeting is provided by the way in which, in many countries, the bulk of the government funding received by public universities is based on their educational outputs. Governments give the universities, say, $6000 per year for each full-time undergraduate medical student, $4000 per year for science or engineering students, $2000 per year for humanities and law students, and so on. The different amounts set for different categories of student reflect cost differences (you need more equipment and facilities to teach science students than to teach humanities students).

Funding based on unit costs creates a powerful link between performance – specifically, the quantity of output delivered – and funds provided. It is broadly equivalent to setting explicit output targets calibrated to the funding level. Unit cost-based funding also has other advantages, including placing different service provider units (in this example, different universities) on an equal footing. It’s therefore easy to see the attractions of this approach.

Funding based on unit costs has been applied to a range of other government services in many parts of the world. In Scandinavia and elsewhere, government schools are funded on this basis. So-called “performance-based funding” of primary health services is largely based on this principle. Many countries also use variants of the unit cost funding model – based on output categories known as “diagnostic-related groups” – in the funding of public hospitals. All these types of arrangements aim to harness the power of a market-like mechanism to promote better government performance.

It is hardly surprising, then, that there are people who think that this is an approach that should be applied across the entirety of government, so as to serve as the basis for the formulation of the whole government budget. Closely related to this is the oft-repeated assertion that, to be effective, performance budgeting must be based on unit costs – as one scribbler recently put it, that “true performance budgeting requires the determination of the unit cost of delivering outputs.”

This makes it important to think about the conditions under which output unit cost budgeting can work and, conversely, those where it will not.

There are two characteristics shared by the public services which are most amenable to funding based on unit costs, both of which are highlighted by the university funding example.

The first characteristic is that each of the output categories used to set specify funding levels is relatively “standardized.” In our university funding example, this means that the activities and inputs required to teach, say, an engineering student are broadly the same for each and every engineering student, and are very similar across different universities. This makes funding based on the standard amount of $4000 per full-time student year workable.

The second characteristic that facilitates unit cost budgeting is that it is possible for the government to decide in advance how many units of output it will fund each financial year – in this case, how many science, engineering, humanities and other students it will finance across the entire government university system. Deciding the output quantity makes it possible for the government to budget based on unit costs while capping its spending on the outputs concerned. This is of great importance for government’s control over aggregate expenditure and, therefore, its ability to achieve its deficit and other fiscal targets.

When we think about the workability of unit cost budgeting in other areas of government, we need to keep these two characteristics – standardization and output quantity control – in mind. Their importance will be highlighted in the next two blog pieces.

But before turning to those matters, we need to note that real-world unit cost funding arrangements are in practice more complex – in some cases, much more complex – than the highly simplified example presented above. As we will discuss in the fourth and final piece in this series, these complexities are one of the reasons why this is a tool which is only suitable for selective application.

Tweet

The Illusion of Outcome-Based Budgeting

The notion that performance budgeting means budgeting based on outcomes is surprisingly widespread. It is, however, a misconception – a misconception which has led to confusion and missteps in PFM reform in some countries, particularly in the developing world.

Outcome-based budgeting refers to the idea that governments should budget by deciding the outcomes they wish to achieve, and then allocating money in the budget to each of those outcomes. The budget would then be comprised wholly, or principally, of appropriations to outcomes.

The idea of a budget structured to focus directly on outcomes is appealing. It is, however, based on a misunderstanding of how budgeting works. Budgets are spent by government entities to produce outputs*, not outcomes. The intention is, of course, that the outputs will achieve outcomes. But outcomes are not in any meaningful sense “produced” by government. When we focus on what the budget actually funds – outputs – we see that, in many cases, outputs contribute to multiple outcomes. For example, school education – an output – contributes (or should contribute) to outcomes including economic development, social cohesion, gender equity, and poverty reduction. When, as in this case, a single output contributes to multiple outcomes, an outcomes-based budget becomes unworkable because it is not possible to translate budgetary allocations to outcomes into allocations for outputs and the entities that produce them. There is, for example, no conceivable methodology which would permit us to designate percentages of school education expenditure which contribute to each of its intended outcomes – that is, which would enable us to say that 40 percent of school education expenditure contributes to economic development, 20 percent contributes to social cohesion, and so forth. This means that there is no logical basis upon which to derive the funding of school education from budget appropriations for economic development, social cohesion, and gender equity.

Consider the further example of reduced maternal and neonatal mortality and morbidity. This is an outcome which is an important priority for many countries, particularly in the developing world. It would be wonderful to be able to budget by identifying priority outcomes such as this, and then deciding how much money to spend on each of them. But the appeal of the idea fades when we think about the practicalities of the matter. It is certainly possible to budget specifically for the outcome of improved maternal and neonatal health to the extent that it involves allocating money to services (outputs) which are uniquely devoted to achieving this outcome – such as maternal health clinics. But that doesn’t take us very far, because most of the government outputs which are crucial to improving maternal and neonatal health are not uniquely targeted at mothers and babies, nor intended to contribute only to the outcome of maternal/neonatal health. To start with, many of the treatments delivered to mothers and babies are provided by general-purpose clinics and hospitals, which are funded with budget allocations designed to cover broader patient populations. It would be inappropriate to pre-specify the percentage of these clinics’ and hospitals’ budgets to be spent on the treatment of mothers and babies because these entities need to have the flexibility to respond to shifts in the demand for their services. Turning to the realm of preventative health, a crucial part of the fight to improve maternal and neonatal health lies in areas such as reducing smoking, improving sanitary habits, and encouraging healthy eating. But most of government’s preventative health efforts in these areas – outputs such as public awareness media campaigns and information-provision – are not targeted exclusively at women of child-bearing age, but at broader population groups.

The problem extends further than this because health services represent only one dimension of the effort required to improve maternal and neonatal health. At least as important are government expenditures designed to reduce poverty, supply clean water, raise female education levels, tackle the problem of child marriages, and increase the economic independence of women. But none of the outputs that government entities deliver in these areas are intended to contribute exclusively to the reduction of maternal and neonatal mortality and morbidity. They all work to achieve other, broader, outcomes. Clean water supplies, for example, improve the health of the public as a whole, and there is no conceivable methodology for estimating the percentage of the expenditure on improving water quality that serves the specific outcome of improved maternal and neonatal health.

All this means is that it is not possible to operate a budgeting system based on allocations to outcomes. Outcome-based budgeting does not, therefore, represent a feasible model of performance budgeting.

This is why the most widespread version of performance budgeting – program budgeting – is a system of budgeting based on outputs and outcomes, rather than outcomes alone. In a program budgeting system, programs – properly understood – represent expenditure on groups of outputs which share common outcomes and other common characteristics (such as a specific client group). The “school education” program, for example, represents a budget allocation to finance the provision of educational services (outputs) provided to school-aged children, the intended outcomes of which include higher literacy, numeracy, economic development, and enhanced social cohesion. The “criminal policing” program is an allocation to the police ministry for policing services (outputs), the intended outcomes of which include crime reduction. Neither of these programs represents a budget appropriation for outcomes such as economic development, social cohesion, or reduced crime.

Contributing to confusion on this matter is the fact that certain countries have chosen to use labels such as “outcome-oriented budgeting” to describe their performance budgeting systems. One should not, however, be misled by this terminology. Such countries do not purport to prepare and execute their budgets based on outcomes alone. Their branding choice reflects, instead, a wish to emphasize the crucial role that outcomes should play in performance budgeting. The reality remains that any good performance budgeting system links funding to both outputs and outcomes

As mentioned at the outset, the flawed notion of outcome-based budgeting has created serious budgeting difficulties in some countries. It has, for example, contributed to the widespread developing-world problem of mismatch between the national plan and the budget. “Mismatch” means that the plan sets priorities, and identifies funds to be directed to these priorities. However, these supposed funding allocations to plan priorities have little relationship to budget appropriations. There are multiple reasons for this mismatch, of which the most important is that most national plans are financially unrealistic. However, the problem is aggravated when, as is often the case, national plans are in large measure outcome-based, and purport to guide the budget on the spending required to achieve outcomes such as private sector development, climate change abatement/adaptation, and the reduction of violence against women. The fact that these types of outcomes cannot be uniquely mapped to outputs creates an insurmountable obstacle in attempting to translate the plan’s financial allocations into workable budget appropriations. This is the sort of problem that arises when planning ministries, which by their nature don’t understand how budgeting works, try – in effect – to take over budgeting by producing plans which purport to be blueprints for resource allocation.

Implementing performance budgeting successfully is not an easy or simple matter. But it can be made much less difficult by avoiding basic conceptual errors like the notion of outcome-based budgeting.

*More precisely, outputs and transfers. “Outputs” is used here as shorthand for both.

Tweet

Making Spending Review Fit for Purpose

Spending review, as I wrote recently, will be needed more than ever in the post-pandemic era. Surely, then, it is encouraging that there has over the past decade been a major expansion in the number of countries practicing spending review?

Well, yes, but there is a problem. The problem is that in some countries, the spending review mechanisms which have been created in recent years are not fit-for-purpose for the post-pandemic era. This is particularly true in certain European Union countries.

It is the very difficult medium and long-term fiscal prospects facing many governments which make spending review essential in the coming years. Spending review has a vital role to play in helping to find fiscal space for additional spending in areas including health, climate change, long-term care and, in some countries, infrastructure. In the many nations with unsustainably high levels of debt, spending review will also have an important role to play in fiscal consolidation – that is, in helping to reduce structural budget deficits and bring down debt.

Finding additional fiscal space and supporting fiscal consolidation require spending review in the classic sense – namely, the systematic review of baseline expenditure to identify savings measures. The focus of such spending review, to put it bluntly, needs to be on making spending cuts.

The trouble is that the more popular spending review has become, the more the term has been de-linked from its original focus on the search for savings measures. It has come to be used to refer to any type of expenditure analysis aimed at improving the quality of public expenditure, even if it has little or nothing to do with the search for savings. Spending review has come to be indistinguishable from other forms of what might be called performance review, such as program evaluation and performance auditing.

Spain is an example. Since so-called spending review was established in Spain in 2016, many reviews have been conducted, but only a mere handful have proposed savings measures. The independent body which has been mandated by the government to carry out “spending reviews” explicitly defines its objective as improving the quality of government expenditure, and is on record as denying any focus on cutting spending.

How is it that the meaning of spending review has morphed in this way in such countries? Much of it can be traced to the anti-austerity backlash after the global financial crisis (GFC). Since 2016, the European Commission has been encouraging – indeed, pressuring – member countries to establish spending review systems. However, many governments – particularly in countries that went through Commission-supervised fiscal adjustment programs after the GFC – want to distance themselves from “austerity” and are unwilling to do anything which might make them appear to be enthusiastic about cutting spending.* They have, however, been prepared to keep the Commission happy by establishing spending review systems with a more uncontroversial focus on improving the quality of public spending. Commission officials have proven willing to go along with this, and have even adjusted their definition of spending review accordingly.**

Don’t get me wrong. I think it’s great to have performance review aimed at improving the quality of public expenditure. It is particularly welcome in countries where the public sector has had, historically, a very weak performance focus. The Spanish “spending review” system is excellent if viewed as an instrument for improving the quality of public expenditure. The independent body which runs the system – AIReF – is doing an outstanding job with the mandate it has been given.

The problem is that this leaves a yawning gap in the place where true spending review should have been. Countries like Spain are not building systems that are well suited to the task of identifying savings measures, despite the fact that this is what ministries of finance and governments will desperately need in the coming years.

There is a need for debate about how best to solve this problem. Governments need both broad performance review and savings-oriented spending review. A crucial question that must be addressed squarely is whether it is possible to package both of these into a single performance review system – in other words, to make the search for savings measures one component of a broad performance review mechanism. I am inclined to think that the answer to this question is “no”. There is historical experience that suggests that if the search for savings options is merely part of a broader performance improvement mandate, it will tend to get lost. There is no surer way of overloading spending review working groups – which are typically required to complete their work in periods of several months – than by asking them not only to identify savings options, but also to search for other possible ways of improving policy, management and the quality of spending generally. If, moreover, review aimed at identifying savings to work is going to work well, it needs to be run by the ministry of finance, rather than by an independent body outside executive government, or some other ministry.

This leads me to believe that the only way of ensuring the effectiveness of true spending review is to maintain, under the direct control of the ministry of finance, a distinct spending review mechanism the primary focus of which is explicitly on the search for savings measures – like the outstandingly-successful Danish spending review system. This, I believe, is the best way to ensure that governments have spending review systems that are tailored to the needs of budget preparation – which is what they are going to need in the years to come.

———————————-

* As an EC official closely involved in the delivery of technical assistance to member countries to build spending review systems has observed “Member States might feel pressured to undertake a spending review to show progress towards a CSR [country specific recommendation] but they might be also mindful of avoiding being perceived as cutting services for their citizens. … The first challenge is often the resistance generated by the common perception of the spending review as a tool to cut the budget. This ‘stigma’ is understandable and it is clearly a legacy of the 2008-2011 economic crisis” (Ercoli, 2020: 51).

** So that it now incorporates reviews aimed “at increasing the value delivered for public money spent by optimising the public funding-impact ratio” (Bova and Ciobanu, 2020: 9).

Tweet

Don’t Confuse Indicators and Targets

One of the problems which bedevils performance budgeting – and performance management more generally – is the failure of some to clearly distinguish between performance indicators and performance targets, and to properly understand the roles of each.

Take the often-repeated mantra that “performance indicators must be SMART.” I’d love to have a dollar for every time I’ve heard that one. It is, however, flat wrong. It is performance targets that should be SMART, not performance indicators.

A performance indicator is simply a measure of something. An example is a literacy indicator such as “level 2 literacy”, which measures the ability to read and write at a basic level. (In the US, for example, approximately one-third of the population have level 2 literacy.) A performance target, by contrast, is a numerical objective for the improvement of performance. If a government declares that it aims to raise level 2 literacy in the general population to 65 percent by 2030, it is setting a performance target. Performance targets are set on the basis of performance indicators, but indicators are not targets.

The most widely-used version of the SMART criteria is:

S = Specific

M = Measurable

A = Achievable*

R = Relevant

T = Time-Bound

Some of these criteria are relevant for indicators. However, the “achievable” criteria is not: it is relevant only to targets. For example, one can and should require that the target of raising level 2 literacy to 65 percent by 2030 be achievable. But it makes no sense to ask whether the current rate of level 2 literacy is “achievable.” A measure is just a measure.

Depending on how they are defined, some of the other criteria may also pertain only to targets. For example, many understand the “time-bound” criteria to mean that an explicit end-date should be set for the achievement of targets – such as the 2030 date in the literacy target above.

The failure to properly appreciate the difference between indicators and targets sometimes manifests itself in even cruder ways. I have, for example, seen many examples internationally of budget documents with tables of “performance indicators” many of which are in fact performance targets – and vice versa.

Even amongst those who are clear about the difference between indicators and targets, one often finds a failure to properly appreciate the role of each. An example is the belief, on the part of some, that for every performance indicator, a target should be set. Certain governments have gone as far as to enshrine this as a rule in their performance management systems. There are, however, many important indicators for which it is inappropriate to set targets because the government simply does not have sufficient control over what the indicator is measuring. This is particularly true of “higher level” outcome indicators. For example, while no one questions the need to estimate and report life expectancy at birth, no practical purpose is served by setting targets such as raising life expectancy at birth by 1.5 years over the next decade. If the government is going to set targets, it makes far more sense to set targets for things over which it has more control, such as the literacy rate of the school-age population. This means targets for lower-level outcomes, output quality, and other similarly manageable variables.

Distinguishing clearly between indicators and targets, and properly understanding the role of each, is a vital element in good performance budgeting and management.

*Or, equivalently, “attainable.”

Tweet

Non-Performing Performance Indicators

It is shocking that, in quite a few countries which embarked on performance budgeting reforms years ago, performance indicators remain awful. Performance indicators are, after all, a basic building block of any performance budgeting system. Yet OECD surveys – most recently in 2018 – have repeatedly shown that in half of member countries, available performance indicators are largely irrelevant for budgetary decision-making. My experience in many developing countries tells me that the situation there is often worse. What is the problem?

The explanation is not that it can’t be done. There are quite a few countries which have, as part of their performance budgeting systems, developed excellent suites of performance indicators. France, Australia, South Africa and New Zealand are examples.

There is an old joke about a drunken man intently searching for his lost car keys on the footpath under a street lamppost. A police officer comes along and, after helping him search for a while, asks him whether he is sure that that is where he lost his keys. The man replies that, no, he lost them somewhere on the other side of the street, but that he is searching under the lamppost because the light is so much better there.

In quite a few countries, it has been exactly like that with the development of performance indicators. Ministries chose performance indicators based on the ready availability of the data rather than the relevance of the indicators concerned. The result is that the budget documents are stuffed with indicators such as the average time taken to fill vacant staff positions, average class sizes, and the number of business licenses issued – because these are things that are typically measured and managed as a routine part of day-to-day administration. (This means, to put it in technical language, measures of inputs, activities (i.e. work processes), and output quantity (i.e. volume of services delivered).) Virtually absent are indicators which are crucially important performance budgeting purposes – namely outcome (effectiveness), quality, and efficiency indicators.

I have seen this mechanism at work in many countries which are in the early days of developing performance budgeting. When undertaking the process of selecting performance indicators for programs, the focus is entirely upon indicators which be developed most easily. Because there is little data on outcomes, quality, and efficiency, few if any of these types of indicators find their way onto paper.

To avoid this, three principles should be applied in the development of indicators for performance budgeting.

The first is that the aim should be to develop, for each program, four types of indicator – effectiveness (outcome), output quantity, quality, and efficiency.*

The second is that important types of indicator should not be omitted simply because no data is available. If, for example, there is no outcome data, the government should nevertheless define the outcome indicators which it needs to develop – be they school student literacy levels, population vaccination rates, or air quality indexes. These indicators should then be included in the list of program indicators in the budget documents with an asterisk indicating that they are under development and will be reported in future.

The third and final principle is that indicators of inputs and activities are, generally speaking, inappropriate for performance budgeting purposes. Knowing how long it takes to fill vacant staff positions is important for internal management purposes, but has no place in a performance budget.

To ensure that these principles are adhered to, the ministry of finance should set clear guidelines for indicator selection and should enforce them, including by reviewing and approving all program indicators proposed by spending ministries.

Developing performance indicators that are truly helpful for budget decision-making is something which takes time and effort. There are no shortcuts here. Using only indicators for which the data happens to be on hand simply guarantees failure.

————————————————————————————————————

* Approximately speaking. There are a few “wrinkles” to this rule – for example, that it does not apply to “administration” programs. For more detail, see my 2011 performance budgeting manual.

Tweet

Population Aging and Public Expenditure: Lessons from Japan

I’ve commented previously on the widespread tendency to exaggerate the impact of population ageing on public expenditure. Sure, it is an important source of upward spending pressure, particularly on long-term care and pensions. However, by far the biggest single area of spending pressure in coming decades will be health, and population ageing plays only a distinctly secondary role in pushing up health spending. The most important driver of rising health expenditure is technological progress – particularly, the arrival of new and better treatments – as I explain in my book Bigger Government.

If you find this hard to believe, you may find it interesting to look at Japanese experience (something which I’ve just done while preparing the introduction of the forthcoming Japanese edition of my book). Over the past fifty years, population ageing in Japan has been dramatically faster than in other advanced countries, and Japan has today a much higher percentage of elderly (65 years and older) and very elderly (80 years and older) people than its peers. However, despite this, total health expenditure/GDP in Japan is today only approximately average for advanced countries. Not only that, but the long-term growth of health expenditure has been quite similar to countries where population ageing has been much, much slower than in Japan.

We can, for example, compare Japan with France and the United Kingdom, both of which have today very similar levels of total health expenditure/GDP to Japan. Over the period 1970-2015, the percentage of the population aged 80 years or older increased by a massive 877 percent in Japan (to 7.6 percent), against “only” 216 percent in the UK (to 4.9 percent) and 259 percent in France (to 5.9 percent). For the 65-plus age group, the increases were 378 percent in Japan (to 26 percent), 138 percent in the United Kingdom (to 18 percent) and 147 percent in France (to 18.9 percent). Despite this, total health expenditure/GDP over that period grew by quite similar amounts in all three countries. (OECD statistics show that between 1970-2018, total health expenditure/GDP grew by 248 percent in Japan, 244 percent in the United Kingdom and 215 percent in France.)

It is simply amazing that in Japan, there is a widespread habit of labeling health expenditure as “age-related spending,” giving a totally misleading impression of the dynamics involved. The OECD and IMF have, unfortunately, adopted the same terminology in discussing the Japan’s fiscal sustainability problem. But those who think that population ageing is the biggest factor driving health spending up need to ask themselves two simple questions. Firstly, if they are right, why is it that Japanese health expenditure hasn’t grown much faster in past decades, in line with the country’s exceptionally rapid pace of population ageing? Secondly, why is it that Japanese health expenditure is not today by far the highest in the advanced world?

Fiscal economists need to abandon the tendency to treat rising health spending as predominantly a problem of demographics. No serious health economist thinks this is the case.

At technical level, fiscal economists in Japan and elsewhere need, as part of a rethink on this important matter, to abandon their habit of forecasting the future growth of health expenditure using the demonstrably erroneous assumption that the so-called “age profile” of health spending will remain constant in future (see my book).

Tweet

Debt, not Net Worth, is what matters

With government debt levels now at exceptionally high levels, many economists and policymakers are reflecting deeply on the challenge of how debt can be reduced to fiscally sustainable levels over the longer term. A few voices have, however, been raised to argue that we should stop worrying about debt and shift our focus to a completely different fiscal measure – net worth. They claim that net worth is a superior measure of fiscal sustainability, and that the focus of long-term fiscal policy should therefore be on increasing net worth rather than on reducing debt.

This view – what we might call the “net worth proposition” – is seriously misguided. What follows explains why.

Fiscal sustainability refers to the capacity of governments to meet their debt and other financial liabilities, without being forced to take sudden damaging fiscal adjustment measures under crisis circumstances (e.g. when faced with a dramatic loss of financial markets confidence in the country’s government debt). Economists generally consider debt/GDP to be the best summary measure of fiscal sustainability for the simple reason that the higher the debt/GDP ratio, the greater the risk of running into trouble. Looking at the current situation it is obvious that, notwithstanding the fact that the interest rates are at present exceptionally low, it is precisely the governments with the highest debt/GDP ratios which are at greatest danger of running into severe budgetary difficulties when, at whatever point in the future it may be, they are confronted with higher interest rate or stubbornly slow GDP growth. It is also the most highly-indebted governments which will have the strongest incentives to pressure their central banks to indefinitely maintain the current currently exceptionally supportive emergency monetary policies – notwithstanding that this would ultimately result in rampant inflation.

It follows that, with so many governments in debt up to their eyeballs, the fiscal sustainability imperative – once post-pandemic economic recovery is assured – will be to gradually reduce debt/GDP. How fast, how soon, and by what means debt should be reduced: these are important questions which will continue to be debated. But surely the long-term need to reduce debt is clear?

Well, apparently not to everyone. By suggesting that we should focus on net worth rather than debt, advocates of the net worth proposition are signaling that they disagree that debt reduction will be required in the post-pandemic era.

Net worth (NW) is the value of government assets minus its debt (more precisely, minus financial liabilities). For example, a government with debt of 120% of GDP and assets with a balance sheet value of 90% of GDP would have a negative net worth/GDP of 30%. So to argue that it is net worth rather than debt which is important is the same as asserting that, from the fiscal sustainability perspective, assets fully offset – i.e. effectively neutralize – debt.

The proponents of the net worth proposition are, in effect, claiming that increasing the stock of government assets is just as good a way of improving fiscal sustainability as is reducing debt. Their reasoning also implies that increases in debt/GDP do not adversely impact fiscal sustainability when the money is invested in acquiring assets. These claims might be music to the ears of some of the governments which are embarking on big post-pandemic public expenditure programs and which may well appreciate being told that they needn’t worry about reducing structural budget deficits and debt. Unfortunately, however, they are claims that do not stand up to even superficial analysis.

Consider, firstly, the notion that increasing the stock of government assets is just as good a way of improving fiscal sustainability as is reducing debt. To see what this would mean in practice, let’s consider a hypothetical example. Suppose that, as one element of a long-term strategy to improve fiscal sustainability, the French government plans in several year’s time to levy a one-off wealth tax surcharge yielding €18 billion, and to use the proceeds to repay some of the country’s astronomic pile of public debt. However, inspired by the net worth proposition, certain government advisors argue that the money could just as effectively be used on the asset side of the balance sheet. Noting the increasingly challenging international security environment, an option they recommend is to add an extra €18 billion to the stock of military equipment – the balance sheet value of which would thereby be expanded by approximately one-fifth. The addition of these extra assets to the government’s balance sheet would have precisely the same impact on net worth as the repayment of debt. According to the net worth proposition, then, this option would be just as good a way of improving fiscal sustainability as is the repayment of debt.

Consider, secondly, the claim that extra debt is fine as long as it is spent on assets. In the context of our hypothetical example, this would translate into a belief that the sustainability of French government finances would not be adversely affected by a decision to fund the extra €18 billion in military equipment entirely through additional debt rather than through the wealth tax surcharge. After all, the extra debt would, supposedly, be “offset'” by the additional assets. Indeed, if we took this line of reasoning a little further, we could argue there would be no problem at all if France took on, say, another €90 billion in debt and doubled its stock of military equipment.

It should be clear that this sort of thinking is absurd. Could anyone really believe that the possession of more tanks, submarines or nuclear missiles improves fiscal sustainability? Given that military equipment yields no revenue whatsoever, and for the most part would never be sold off even during a severe financial crisis, how could it possibly be suggested that these assets offset debt?

The example illustrates a general point. This is that assets on the government balance sheet can only legitimately be treated as offsets against debt if, and to the extent that, those assets are recorded in the balance sheet at what we might call their “financial value” – meaning what they are worth in terms of either revenue-generation or practically-realizable resale value. Only insofar as assets have financial value are they capable of yielding cash which could be used to meet government liabilities. But most assets* on the government balance sheet are not valued on this basis, but rather on some measure of their cost**. The €90 billion valuation of the military equipment in the France’s government balance sheet is based on what the government paid for the equipment, and is vastly in excess of its realistic financial value. The same gulf between balance-sheet value and financial value applies to many other government assets: such as, for example, schools and urban roads, both of which yield little or no revenue and could not, in most cases, conceivably be sold to pay off debt*** – but which nonetheless have very considerable balance-sheet valuations.

The total balance sheet valuation of government assets is, as a consequence, many times larger than those assets’ financial value. This is not because of the valuation practices built into government accounting are bad. Those valuation practices simply reflect the fact that for government – unlike the private sector – many assets are held for their social benefits rather than financial returns. This fundamentally important reality makes the government’s balance sheet conceptually completely different to the balance sheet of a private corporation.

It follows that, for government, net worth is a very poor indicator of fiscal sustainability.

I’m not suggesting that assets are entirely irrelevant to fiscal sustainability. There is a respectable argument for treating the government’s financial assets (money in the bank, debts owed to government, share holdings etc) as offsets against debt. This is why economists often use various measures of so-called net debt (debt minus financial assets)****. Non-financial assets can also be relevant, although only to the quite limited extent they have practical financial value – as in the case of surplus land. But to accept that balance sheets provide useful information for fiscal policy makers does not in any way imply that they should stop worrying about debt and focus instead on net worth.

The fact that we are now in the post-“austerity” era should not be allowed to muddy the waters on this important matter. The debate over the respective merits of net worth and debt as fiscal sustainability measures has nothing whatsoever to do with how we feel about the merits of Keynesian fiscal stimulus. One can accept the need for some additional spending over the next couple of years to ensure post-pandemic recovery – even at the expense of some further short-term increase in debt – without in any way endorsing the net worth proposition.

Nor does this debate have anything to do with how we feel about the so-called “golden rule”, which is the proposition that debt can be justified on inter-generational equity grounds if it is used to finance capital expenditure which creates assets which will benefit future generations. Even if one is favorable to the golden rule (as I am), one must recognize that ensuring inter-generational equity does not ensure fiscal sustainability. Fiscal sustainability is of overriding importance, and trumps inter-generational equity. So if a government already has unsustainably high debt, it cannot legitimately use the golden rule to justify further adding to its debt burden for capital expenditure purposes. This makes the golden rule of practical significance only for those – sadly, relatively few – governments which have debt already at sustainable levels.

Debt is not a perfect fiscal sustainability measure, but it is by far and away the best measure that we have. Governments should therefore continue to use debt as a core measure for the specification of fiscal rules and targets. Any suggestion that net worth should replace debt for these purposes should be firmly resisted.

NOTES

(*) Most non-financial assets, to be more precise.

(**) The reality that most general government assets are valued on a cost basis in real-world balance sheets is more clearly recognized in the International Public Sector Accounting Standards Board’s recently-released ED77, with its emphasis on the “current operational value” of assets which are held for the services they provide rather than the financial flows they yield. IPSASB prepared these new standards in response to criticisms from governments that the undue primacy given to the concept of “fair value” in accounting standards. (As an aside, while these proposed new standards are a positive step forward, the economic logic behind their emphasis on the replacement cost of assets is unclear. Real – i.e. inflation-adjusted – historical cost, makes more sense.)

(***) The point is not only, as is often suggested, that these assets are “illiquid” (i.e. difficult to sell and capable of being sold only with considerable time delays). Of much greater importance is the fact that governments would never consider selling many of the assets concerned. School buildings and land would not, for example, be sold however serious a financial crisis might be (with limited exception such as the odd schools in a depopulated area). Governments would invariably prefer default to the sale of such crucial assets.

(****) It can be argued that “net financial worth” – defined as liabilities minus financial assets – is a particularly useful broad debt measure. (Expressed differently, net financial worth = net worth – non-financial assets.)

Tweet

Strengthening New Spending Decision-Making

Disciplined decision-making about new spending is crucial to good budgeting. Yet it is something which is absent, or weak, in many countries. This is particularly true in developing countries, but is by no means confined to them. A key part of the problem is of course the nature of politics, which creates a powerful bias to ad hoc spending decisions designed primarily to buy votes and to respond to the short-term news cycle. Coupled with this is a tendency for new money to be allocated to the ministries which are led by the most powerful ministers, irrespective of the merits of the proposed spending. And governments have, after all, a natural inclination to spend more than is affordable.

This means that, in many countries, a major focus of budget system reform needs to be on building processes and institutions to ensure that any new spending is limited to initiatives which are high-priority, cost-effective and fiscally sustainable. This is particularly important at the present time, when the difficult post-pandemic fiscal position of many government makes it essential to be particularly disciplined about any new spending initiatives. This includes being very careful about what is contained in post-pandemic economic recovery packages.

Reforms to address this challenge should aim to ensure that:

No new spending decision is made without the proposed spending initiative being subject to tough internal critical analysis within government, aimed at rooting out ill-considered proposals.New spending proposals should – with only the most limited exceptions – be considered by government only as part of the normal budget preparation process. Ad hoc “out of cycle” spending proposals should be minimized.All new spending proposals should be reliably costed, and total new spending should be limited to levels consistent with available fiscal space.

The last of these points is a core element of good medium-term budgeting, and is generally well understood. So let’s focus on the other points here.

It is the role of the bureaucracy to subject all new spending proposals to searching critical analysis, and to provide fearless advice to political leaders on their merits (or lack thereof). This is a responsibility which is not borne exclusively to the Ministry of Finance, although the MOF certainly has a pivotal part to play. Other relevant ministries should also routinely view, and comment on, major new spending proposals prior to cabinet or presidential decisions. The president’s or prime minister’s office should work closely with the MOF in scrutinizing proposals from the perspective of whole-of-government priorities and value-for-money. Any spending ministry that wishes to obtain government approval for the substantial new spending proposal should understand in advance that it will need to “run the gauntlet” in order to do so.

For the process of critical internal review of new spending proposals to work properly, there have to be appropriate procedural rules, and these have to be adhered to. These include an adequate minimum time period between a proposal being put forward by spending ministry and the decision by the political leadership on whether to approve it. The idea here is to make it impossible for a new spending proposal to be “sprung on” the cabinet at the last minute by a spending minister as a tactic to prevent proper review and debate on its merits.

In order to facilitate all this, all new spending proposals should be submitted in accordance with standard templates which make it mandatory for the spending ministry to provide relevant key information on the proposal. This means not only financial information, but also information on objectives and “intervention logic” (i.e. why it is considered that the initiative will achieve its intended objectives, and that it is the best way of achieving these objectives). The performance indicators which will be used down the track to help assess the effectiveness of the spending initiative should be specified – this in one way of building a bridge between the new spending decision process and the performance budgeting system.

The quality of the bureaucracy, and its policy analysis skills, is of course fundamental. This is just one of the many reasons why bureaucratic capacity-building is so important in developing countries. The civil service needs to be professional and merit-based. The politicization of the civil service, and the consequent rapid turnover at senior levels when the government changes, is a major obstacle to this in many developing countries. One region where this problem is particularly rampant is Latin America, as emphasized in an important recent OECD/BID report. But this is also a major problem in some advanced countries, the United States being one of the most extreme cases.

The bureaucracy must, however, also be allowed to do its job. If civil servants are largely excluded from decision-making about new spending – with this role being essentially taken over by extra-governmental party-political bodies – the inevitable result will be wasteful and irresponsible spending. This is one of the things which, for example, has gone wrong in South Africa, as emphasized in an impressive recent paper by the country’s former budget director, Michael Sachs.

There will always be some “out of cycle” new spending decisions – the most obvious example being unplanned spending in response to natural disasters. But it is important to keep these to an absolute minimum, so as to ensure that under normal circumstances almost all new spending proposals compete against one another during the budget preparation process so as to ensure the “survival of the fittest” and to help guarantee that new spending does not, in aggregate, exceed available fiscal space. For this purpose, there need to be clearly-defined criteria which specify precisely what limited types of new spending proposals are permitted to be put forward outside the budget cycle.

What about the role of planning in guiding new spending? There can be no question that good planning – at both the ministry and whole-of-government levels – is crucial. Planning is about clarifying objectives and means of achieving them. But it would be a mistake to think that good decision-making about new spending is entirely a question of having good planning. In particular, it would be wrong to believe that all the heavy lifting of decision-making about new spending should occur in the context of the preparation of the plan – and to think that any new spending idea which appears in the plans should therefore automatically be incorporated into the budget. Plans are generally not fiscally realistic, and they rarely reflect definitive political decisions about the allocation of scarce fiscal space. The role of good planning should therefore be to inform budgeting, not to take over budget decision-making.

These are merely a few of the points which are crucial to ensuring good decision-making about new spending. The objective, it should be emphasized, is not be to de-politicize new spending decisions. Decisions on the allocation of limited public resources are, and will always be, inherently political, because they necessarily reflect political values, ideology and objectives. The true objective is to help political leaders direct spending to areas which make sense in policy as well as political terms, while remaining fiscally responsible.

Tweet

Spending Review is Back

Governments will in the coming years need spending review more than ever before. Given the big surge in new spending which is currently underway, this might seem a strange statement. Spending review is, after all, the systematic search for areas where government spending can be cut. Many will argue that the last thing we need to be focusing attention on in the post-pandemic era is spending cuts. Doesn’t everyone agree that the era of austerity is over?

It is quite true that there is wide agreement that stimulus spending has a major role to play in the recovery from the pandemic, and that premature action to eliminate budget deficits would be entirely inappropriate. But that should not blind us to the fact that, viewed in the longer term, the finances of most countries are unsustainable. This was true for many even before the pandemic. It is much more true today. Not only have debt levels increase dramatically, but many countries have taken recent tax and spending measures of a permanent nature which have further worsened their structural budgetary positions. It would be naïve to think that super-low interest rates will continue forever and save us from the consequences of high debt. The blunt reality is that, once economies have recovered, fiscal consolidation will be essential. Structural budget deficits will need to be reduced, and debt gradually brought down.

This does not necessarily mean cutting government spending to below pre-pandemic levels. While this is necessary in some countries, it would be entirely inappropriate in others. Moreover, there are – as I’ve argued in my book Bigger Government – powerful longer-term forces pushing government spending up.

But spending review should not be thought of as simply a tool for cutting aggregate spending. It has another core function – that of increasing the fiscal space available for high-priority new spending. To equate spending review with “austerity” is fundamentally silly.

Some might think that tax increases can save us from the need to make spending cuts. Of course, tax rises have a significant role to play in restoring public finances. This is particularly true in low-tax countries, including the United States and many developing countries where tax collection is poor. But it would be unrealistic to think that the entire weight of fiscal consolidation – plus the financing of high-priority new spending in areas such as the energy transition, health and infrastructure – can be borne by tax rises.

In this context, designing good spending review processes is essential. The lessons about spending review design which I drew in my 2013 paper for the OECD remain as valid as ever. A couple of points are worth underlining:

Spending review must be an integral and ongoing part of the budget process, not just a “one-off” exercise.It is a mistake to think that the focus of spending review should be wholly or even primarily on making efficiency savings – i.e. finding ways of saving money in delivering existing services and achieving outcomes. There is a tendency to exaggerate the magnitude and, above all, the speed with which efficiency savings can be delivered.Spending review must also therefore give great attention to strategic review – that is, to identifying cuts which can be made by scaling back or eliminating government activities in areas which are deemed to be low-priority or to deliver poor value for money.The civil service has a core role to play in spending review. Spending review cannot simply be delegated to outside experts and business leaders, notwithstanding the useful role that they can sometimes play. This makes the capacity of the civil service — its policy and analytic skills — fundamental.Evaluation is not spending review. Evaluation reports should be viewed, rather, as an input — an information source — into spending review. Much spending review can and should be carried out without elaborate evaluations. While it is important for governments to have well-developed evaluation capacity, history teaches us that we must be careful to avoid creating an evaluation industry the output of which is of little use to budgeters and managers (see my 2014 paper). There are some countries (e.g. in Latin America) which have quite elaborate institutionalized evaluation structures, but a poor record in spending review and budget re-allocation.Tweet

Accrual Budgeting: Think Carefully Before Jumping

[Warning: this is a technical blogpiece for public financial management specialists]

Is moving to a accrual budgeting a good idea? This is a question I am asked from time to time. It is a controversial matter which needs to be carefully considered by any country contemplating making this move. Accrual budgeting has benefits, but also brings added complexity – and the two need to be weighed against one another. It is certainly not the case that the adoption of accrual accounting necessarily means that a country should also adopt accrual budgeting. Combining accrual accounting with some form of cash budgeting remains a perfectly acceptable option. It is also not the case that, as is sometimes suggested, accrual budgeting is required if a country is to have an effective performance budgeting system.

Moving to an accrual budgeting system requires a number of key technical design decisions. There is no one unique form of accrual budgeting which constitutes “best practice” which may be simply adopted wholesale. Amongst these design decisions are the following:

How will capital budgeting work?

The defining feature of accrual budgeting is that ministries and agencies are given budgets for their accrual expenses, as opposed to their cash payments. But whatever the budgeting system, it is crucial to maintain control over capital expenditure, and capital expenditure is not an expense. So it is essential to combine budget allocations for expenses with some form of budget allocations covering capital expenditure – broadly in the same manner that ministries are given capital expenditure allocations under cash budgeting systems. The crucial question then is what concept of capital expenditure is used for this purpose, including whether it is modified in accord with accrual concepts. An associated question is whether or not to give the depreciation component of expenses allocations a role in the budgeting system.

How will the accrual budgeting system be integrated with access to cash for payments?

Under an accrual budgeting system, ministries and agencies will receive expenses allocations which in part cover expenses incurred this year which are not paid until next year or subsequent years (e.g. some bills payable or employee leave entitlements). The question then is how to regulate their access to the cash required to make such payments, in such a manner that they do not spend – in cash terms – significantly more than their past expenses budgets entitle them to spend. The appropriate answer to this question will not necessarily be the same for all countries.

Closely associated with this is a major transitional issue which arises when moving from a cash budgeting system to an accrual budgeting system. In the year this transition is made, ministries will find that they need to make many payments arising from past expenses which are not part of the transitional year’s expenses budget. Arrangements need to be made for them to receive an appropriate transitional cash allocation.

How will commitment control work?

Commitments control is an essential part of any budgeting system. This is just as true in accrual budgeting system as in a cash budgeting system, because there are many types of spending commitment which only result in expenses in future years. The signature of a multi-year contract for an infrastructure project is an example. This point is not always understood – there are some who seem to think that the problem of commitment control disappears if one moves to an accrual budgeting system.

These are merely a sample of the range of major technical questions which arise if it is decided to move to an accrual budgeting system. It points to the need not only to reflect carefully before deciding to make the move but also – once having decided to move to accrual budgeting – to put appropriate effort into the design and sequencing of what is a highly demanding reform.

Some of these issues are discussed further in past writings of mine on this topic, which may be found on my website.

Tweet