When windmills tilt: The FRBM debate

  • Blog Post Date 08 May, 2017
  • Perspectives
  • Print Page
Author Image

Pronab Sen

Chair, Standing Committee on Statistics

In this article, Dr Pronab Sen presents his views on the ongoing debate on the Fiscal Responsibility and Budget Management Act.

Recently, the Indian Ministry of Finance released the Report of the FRBM (Fiscal Responsibility and Budget Management Act) Review Committee1, whose volume 1 is somewhat grandiosely entitled: Responsible Growth: A debt and fiscal policy framework for 21st century India. This is a massive piece of work running into four volumes, and I have no intention of getting into its details. The purpose of this article is more limited. The most fascinating part of this Report is the Note of Dissent by one of the members of the Committee2(Annex–V of Vol. 1 of the Report) and the Rejoinder of the Committee to the Note of Dissent (Annex–VI). Usually such dissent notes are of little consequence, except for the political opposition or for the academically inclined. But this case is different: the dissenter is a very influential official of the Ministry of Finance, which set up the Committee in the first place, and which is the ultimate decision-maker on the Committee’s Report.

It may, therefore, be of some interest to examine the extent to which the dissent note diverges from the Committee’s recommendations, and the relative merits of the divergences. At the heart of the disagreement is a rather fundamental issue: should the target of fiscal correction be government debt, as recommended by the Committee, or the primary deficit3, as advocated by Subramanian. In his dissent, Subramanian argues that India should prepare for the inevitable day of reckoning when the growth rate of nominal GDP (gross domestic product) will become roughly equal to the rate of interest. In such a situation, debt sustainability requires that the primary deficit should be zero or less.

Fiscal deficit as “operational” target

Subramanian objects strenuously to the “serpentine path for the Centre’s fiscal deficit”, which the Committee has recommended. This path involves a large 0.5 percentage point reduction in the first year, followed by no change for two years, then a gradual reduction of about 0.2 percentage points annually for the next three years. Instead he proposes a “gentle glide path” of 0.2 percentage points per year decline in the primary deficit, which would reduce the fiscal deficit to 2% of GDP as against 2.5% over the same five-year period. It is not clear at all why he couldn’t simply have recommended an equally gentle glide path of 0.22 percentage points per year for the fiscal deficit, which would have yielded the Committee’s target of 2.5% in the terminal year.

In its spirited response, the Committee makes much of the fact that a “declining primary deficit … is neither necessary nor sufficient for debt ratios to decline”. This is absolutely correct. What they conveniently forget to mention is that this statement is equally true of the fiscal deficit as well. This is not duplicity, but either just a typo or a genuine algebraic error, which can happen to anyone4.

Be that as it may, the Committee is extraordinarily coy about why it recommended the “serpentine path” and not a gentle glide path to begin with. I suppose it is embarrassing to cite political considerations in what purports to be a technical document. Nobody is really fooled – clearly the path has been dictated by the fact that the next general elections are in 2019, and some flexibility had to be provided for the government of the day. But the 2017-18 budget is already history, and no government in its right mind is going to do a 0.5 percentage point reduction in the fiscal deficit ratio in what is effectively its election budget5.

So what can we expect? By the Committee’s recommendation, the Centre’s fiscal deficit would be 3% in 2018-19 and 2019-20. If the 2022-23 fiscal deficit target of 2.5% is accepted and a glide path is followed, then the figures will be 3.3% and 3.1% respectively, which provide a whole lot more flexibility. I would bet on the glide path.

However, it is surprising that Subramanian does not point out that while his proposal in all probability will meet the Committee’s objective of reducing the debt ratio, the Committee’s proposal, on the other hand, in all probability will eventually lead to a rise in the primary deficit, and thereby enhance the potential vulnerability of the Indian economy. This arises from the fact that the time paths of the fiscal and primary deficits are related. On the basis of this relationship, it can be shown that so long as the annual reduction in the Centre’s fiscal deficit ratio is more than 0.1 percentage point (which it is) the primary deficit will also reduce. But when the fiscal deficit is stabilised at the recommended 2.5% of GDP, the primary deficit will start rising and will continue to do so as long as the growth rate of nominal GDP stays above 7% per annum7. I do not think any one expects India’s GDP growth to dip below this figure on a sustained basis in the foreseeable future; and if it does, the FRBM will have to be reconsidered anyway.

I believe that Subramanian wins this round on points8.

The government debt/GDP ratio

Subramanian’s main objection to the Report is its use of the debt/GDP ratio as the “anchor” for fiscal policy. He rightly points out that there is no reason to believe that India’s current debt ratio is either undesirable or unsustainable. He further draws attention to the fact that in the entire literature on this subject, nobody has been able to derive any estimate of an ‘optimum’ level of the debt ratio. The closest has been in a paper by Reinhart and Rogoff (2010)9, which set the debt threshold at 90% of GDP. Unfortunately, this paper has been thoroughly discredit10, which now leaves the cupboard entirely bare.

Having said all this, however, he complicates things by asserting that what matters is not the level of debt, but the direction the debt is heading – downwards, good; upwards, bad (what about constant?)11. I do not know whether he actually believes in this, but he certainly needs it to justify his position of reducing the primary deficit until it turns positive. The problem with this argument is that if every year investors want to see the debt ratio declining in order to buy fresh public debt, eventually the debt ratio will have to be driven down to nearly zero. Moreover, as debt reduces, so will the fiscal deficit. Simulations using a variety of alternative assumptions suggest that Subramanian’s target of a small positive primary balance will lead to Centre’s debt becoming 20% of GDP in 15 years – less than half the current level of 47% - and continuing to decline thereafter13. Over the same period, the fiscal deficit will go down to a mere 0.7% of GDP.

The Committee has defended its position by asserting that its target consolidated debt ratio of 60%14 is not a figure drawn out of a hat, but a reasoned estimate based on no less than seven different approaches. Fair enough, but they promptly lose the moral high ground by conceding that this number is not really a target, but a ceiling. The reason they are forced into this concession is that for any GDP growth rate of above 7%, a stable central fiscal deficit of 2.5% will lead to the debt ratio continuing to decrease, much like Subramanian’s primary deficit target, but not to the same extent15. There are only two ways to address this problem: (a) have the courage of conviction to state that once the target central debt ratio of 40% is attained, the fiscal deficit may be allowed to rise16; or (b) estimate the level at which the debt ratio will stabilise at a constant 2.5% fiscal deficit, and ask whether this level is appropriate for the economy.

Option (a) will certainly ruffle the feathers of fiscal hawks, and perhaps international investors, but that is hardly a good reason not to do the right thing17 Option (b) unfortunately runs into an insuperable problem – there is no analysis that I am aware of on what is the minimum public debt that a modern economy needs to function properly, nor even the considerations that should go into such an assessment. All existing work relates to how much public debt is too much or how much fiscal deficit is too much. Nobody seems to have asked the question as to how much public debt or fiscal deficit is too little. This, I believe, is an extremely serious gap in the public finance literature, and I am grateful to this debate for forcing me to think about it.

Why do I believe that this is important? The answer lies in the fact that government securities are the only risk-free interest-earning instruments in any economy, and are therefore central to a wide range of key economic variables and decisions. I have not thought through this as yet, but the level and structure of interest rates, conduct of monetary policy, risk profile of the financial sector, are some instances which immediately spring to mind. These are all systemic issues, and I am sure there will be others of both systemic and sectoral importance. It should be noted that these considerations involve not only the stock of government debt, but also the flow (that is, the fiscal deficit).

If my apprehensions are correct, then any proposal that leads to an uncontrolled reduction in both the debt stock and the fiscal deficit takes us into uncharted waters and should be eschewed until we have some sense of how far they can be allowed to fall without creating systemic turbulence. Subramanian’s primary deficit target does exactly that18.

This round therefore goes to the Committee on substance.

The “escape clause”

A very important innovation of the Report is its introduction of a rule-based deviation from the fiscal path during times of large swings in the economy. The absence of such a provision in the existing FRBM was a serious weakness which resulted in the government of the day ‘pressing the pause button’ on multiple occasions. The consequence of course is that once the pause button is pressed, all restrictions are off and the government can do pretty much as it pleases, including when and how to get back on to the FRBM path.

Subramanian “welcomes” the Committee’s recognition that the FRBM needs an escape clause, but has issues with its formulation. He and the Committee agree that the best approach would be by “cyclically adjusting the deficit targets” and that “calculating such adjustments is not possible at the present time”19. He then proceeds to criticise the Committee’s proposal for triggering the “escape clause” only in exceptional circumstances, that is, when growth diverges by 3 percentage points or more from the average of the latest four quarters. He points out that until the escape clause is triggered, the government’s behaviour is likely to be “dramatically pro-cyclical, aggravating the slumps and booms”. In this he is absolutely correct, but he stops short of providing an alternative formulation.

The Committee’s response is surprisingly meek in that they justify their proposal for allowing deviations only when the conditions are “large” and “rare” on the grounds that it preserves “the sanctity and credibility of the framework recommended by the Committee”. They go on to recommend that this issue “should be looked into seriously by future reviews of the fiscal framework”. Oh, really? What do we do until then?

This round, and therefore the debate, is a draw.

I strongly suspect that by the time they got around to this extremely important issue, the entire Committee, including Subramanian, were suffering from acute mental fatigue – they just wanted to get over with it20.

However, we can’t really leave it at that, because the consequences can be very serious indeed. The starting point for making any progress is the recognition that:

  • Pro-cyclicality is the inevitable outcome of any binding constraint on public borrowing, whether it is through the fiscal deficit or the primary.
  • Centre and states should have different rules.
  • Shocks and cycles are very different, and separate provisions need to be made for each.
  • Return paths to the fiscal rule are sensitive to the nature of the deviation and the escape clause should reflect this.

In what follows, I will confine myself to the fiscal deficit rule recommended by the Committee and stay as true as I can to the substance of the Report.

As a country, we should not be unfamiliar with the pro-cyclicality of binding borrowing limits – most of our stateshave experienced this over the last 12 years. What is interesting is that while all of them have been forced to curtail their expenditures during slumps, many have not fully utilised the available fiscal space during booms. Clearly, fiscal probity is not as rare a virtue as people think. Nevertheless, given the heterogeneity of state-level behaviour and the difficulty of inter-state coordination, it may be wise to not provide an escape clause for states and let the entire burden of adjustment be borne by the Centre, which has many more policy options than do the states21. An immediate implication of this is that the Centre will have to then take into account not only the cyclical dynamics of the broader economy but also the pro-cyclical behaviour of states while framing its counter-cyclical interventions.

As the Committee itself points out, the recommended triggers are large, rare events, that is, shocks, which have three characteristics: (a) asymmetric, in that almost all shocks are negative22; (b) sharp and sudden, with usually slow recovery23; and (c) usually not amenable to monetary policy24. In order to prevent prolonged disruption, therefore, the fiscal response must be equally large and sharp. The Committee’s recommendation of a 0.5 percentage point relaxation of the fiscal deficit is simply not enough25. However it is calculated, the expenditure multiplier26 in India is no more than 3, which means that if a 3 percentage point decline in GDP is to be countered, the fiscal deficit will have to rise by at least 1 percentage point of GDP, and probably significantly more to take into account the contraction in expenditures by states and possible pro-cyclical monetary action27. The reversal of such a deviation will need to depend upon the duration of the shock and pace of recovery.

Cyclical movements, in contrast to shocks, are gradual, usually symmetric, and amenable to monetary correction, which makes them easy to overlook in the short-run. However, they can be as pernicious as shocks, especially if the government as a whole is committed to a pro-cyclical fiscal stance. Not providing for such eventualities can expose the economy to unintended instability, especially since the monetary response would necessarily have to be larger than otherwise. In its response, the Committee has indicated that it has considered deviations of 2 percentage points from the four-quarter average, but has rejected it in favour of the 3 percentage point trigger. I would suggest that consideration may be given to a trigger at 1.5 percentage point deviation in both directions with a flexibility of 0.3 percentage point in the fiscal deficit. This may not entirely take care of the cyclical issue, but would contribute to limiting the potential damage.

Finally, the ability of the government to reverse any deviations from the long-term fiscal path depends crucially on the nature of the expenditure commitments28. The problem occurs when the increase in expenditures is on activities which are difficult to roll back for political reasons. By and large, such activities fall into the category of revenue expenditures, while capital expenditures can be turned on and off with much greater ease. It is, therefore, important to make clear in the escape clause that only the fiscal deficit target is being relaxed, and not the revenue deficit target. The existing formulation misses this point altogether.

In conclusion

While I fully realise that all unsought advice is gratuitous, the open disagreement within the FRBM Committee does call for some input into the final decision that the government will have to take. I offer my advice for whatever it is worth:

  • Accept the 2.5% fiscal deficit operational target for 2022-23 with a clear realisation that it can be relaxed up to 4% without compromising the 40% debt/GDP objective. Retain the limits for states as specified.
  • Provide two escape triggers – one at 1.5% change in the growth rate with a 0.3% flexibility in the fiscal deficit; the other at 3% change in the growth rate with 1.5% fiscal deficit flexibility.
  • Do not permit any flexibility in the revenue deficit target.
  • Immediately set up a committee to examine: (a) how far the Debt/GDP ratio and the fiscal deficit can be allowed to fall without serious systemic repercussions; and (b) the transition path to a zero primary deficit regime.


  1. Popularly known as the N.K. Singh Committee.
  2. Who happens to be Dr. Arvind Subramanian, the Chief Economic Adviser to the Ministry of Finance.
  3. The primary deficit is the fiscal deficit minus interest payments, that is, that part of the deficit which is spent on real goods and services.

Further Reading

  • The Committee specifies the “simplest equation that drives the evolution of debt” as:

    dt+1 – dt = [1/(1 +g)] + fdt+1.

    This is derived from equation 5 in Annex 1 of Chapter 4 of the Report. It is simply wrong, and the correct expression is:

    dt+1 – dt = - dt*[g/(1 +g)] + fdt+1

    What this means is that the fiscal deficit ratio in any given year must be lower than a threshold level given by the initial debt ratio and the growth rate of nominal GDP for the debt/GDP ratio to decline. The threshold at 11.5% GDP growth is 7% and at 7% GDP growth is 4.5%. Since the consolidated fiscal deficit is around 6.3%, no fiscal correction is necessary at the former, and a fiscal deficit reduction of less than 1.8 percentage points is not sufficient at the latter.

  • Unless of course the fiscal year is abruptly changed from April-March to January-December, which seems to be very much on the cards if recent pronouncements are anything to go by.
  • Algebraically, assuming that the interest rate on government debt (r) stays constant, the relationship is:
  • (pdt+1 – pdt) = (fdt+1 – fdt) - r*[fdt – dt*g/(1+g)]

  • The primary deficit ratio will asymptotically converge towards the 2.5% fiscal deficit target from its current level of about 0.8%.
  • Although I do not necessarily agree with him.
  • Reinhart, CM and Kenneth S Rogoff (2010), ‘Growth in a Time of Debt’, NBER Working Paper No. 15639.
  • Herndon, Thomas, Michael Ash and Robert Pollin (2013), “Does high public debt consistently stifle economic growth? A critique of Reinhart and Rogoff”, Cambridge Journal of Economics, December.
  • The argument rests on investor perceptions and their willingness to purchase government securities.
  • Again, asymptotic convergence.
  • In the case of the states, the reduction is less dramatic at 15% as compared to the current 21%.
  • Which is divided up into 40% for the Centre and 20% for the states.
  • Subramanian clearly recognises this problem of inconsistency between the debt ratio target and the fiscal deficit rule and uses it as a debating point.
  • Simulations again suggest that a final resting place of the fiscal deficit between 3.5% and 4% may be appropriate depending upon the assumptions made regarding future growth trajectories.
  • It could also draw snide remarks from Subramanian, since the path would become even more “serpentine”. But would he really object to a nice neat U-shaped path?
  • India will eventually have to move towards a primary balance target, but this is not the time.
  • The term ‘cycle’ in this context relates to reasonably predictable movements in the growth rate, both up and down, over the course of a business cycle. There is a view that India has never had a business cycle, and all variations have been driven by shocks of various kinds. Since predictability is a function of past patterns, we therefore cannot devise a strategy to counteract such swings. In my opinion this view is not correct, and that we have experienced one cycle. However, the length of this cycle (peak-to-peak) has been 10 years as against roughly four years in the developed countries, which means we cannot readily draw upon their experience in framing policy rules.
  • Sadly, this chapter, which is the most original and possibly the most important, is also the weakest in the Report.
  • The Report is not clear about whether flexibility is to be given to states, and this can lead to confusion.
  • There can be positive shocks, such as the dramatic collapse of oil prices in 2015, but these are even rarer.
  • A 3 percentage point decline in growth in one quarter is very sharp.
  • The phrase “pushing on a string” popularised by John Maynard Keynes refers precisely to the ineffectiveness of monetary policy in addressing large negative shocks. Moreover, since ‘inflation targeting’ has become the core mandate of the Reserve Bank of India, large supply shocks which tend to push up inflation may actually lead to monetary contraction, which will simply make matters worse.
  • And Urjit Patel’s recommendation of 0.3 percentage points is simply risible.
  • The expenditure multiplier is a measure of the extent to which the GDP responds to a change in autonomous 2 expenditure. Thus, a multiplier of 3 means that for every Re. 1 increase in government expenditure GDP will rise by Rs. 3.
  • My own sense is that the provision should be as high as 1.5 percentage points.
  • Changes in tax rates, on the other hand, are relatively easier to reverse.
No comments yet
Join the conversation
Captcha Captcha Reload

Comments will be held for moderation. Your contact information will not be made public.

Related content

Sign up to our newsletter