Archives: Publication

The Lead Bank Scheme of the RBI involves a ‘lead bank’ in each district of India that coordinates financial services in the district and a ‘convenor bank’ in each state that monitors these activities of the lead banks in the state. Analysing quarterly district-level data on credit disbursement for 2004-2019 this article finds that in districts for which lead bank and convenor bank belong to the same corporate entity credit disbursement is 21% higher.

The Reserve Bank of India’s (RBI) ‘Lead Bank Scheme’ is among the oldest financial inclusion programmes in the country. Initiated in 1969 the scheme assigns the role of lead bank to a public-sector bank in each district based on the overall capabilities of the bank in the district. The lead bank coordinates with local financial-market agents to identify and address bottlenecks in the provision of financial services and to conduct public outreach programmes such as financial literacy camps in the district. Further another public-sector bank is assigned the role of convenor bank for the state which involves monitoring the performance of lead banks of the state. The chairman of the convenor bank executes this function through quarterly meetings with lead banks. These meetings are regularly attended by senior bureaucrats and politicians of the state and representatives of local press and NGOs (non-governmental organisations). If it is found that the lead bank has failed to meet targets penalties are imposed which may include allotting some funds to (low-yield) Rural Infrastructure Development Fund.

Given this organisational structure of the scheme some districts have their lead and convenor banks belonging to the same corporate entity. What effect does this alignment have on the financial-market outcomes in the district? Theories of organisational economics suggest that lower costs of intra-firm monitoring and punishment lead to higher effort by employees (Williamson 1981 Holmstrom 1999). In this context the lead bank of an ‘aligned’ district should be exerting more effort as they are reviewed by their immediate supervisors. If so one can expect higher credit disbursement in these districts.

For most public-sector banks in India recruitment is conducted only for junior-level employees whereas the senior positions such as for lead bank personnel are filled through promotions based on merit-cum-tenure. Further attrition and termination of employment remains low in these organisations. In such an environment managers subject to review by their immediate superiors are more likely to exert higher effort in the assigned tasks.

If the personnel of the lead bank experience added organisational pressure from their own supervisors if the convenor bank belongs to the same corporate entity then one can expect credit disbursement through branches of lead banks to be higher in ‘aligned’ districts compared to branches of other banks in that district and branches of the same bank in non-aligned districts. However data on branch-district level credit disbursement are not publicly available (see Acharya et al. 2019). Therefore I test the hypothesis using RBI data on quarterly district-level credit disbursement from 2003-04:Q4 (fourth quarter of the financial year) to 2018-19:Q4 (Gupta 2020). Nearly 44% of the districts are aligned. All else being equal aligned districts should have higher credit disbursement as monitoring by one’s own chairman should imply higher effort by lead bank personnel in aligned districts.

District ‘alignment’ and credit disbursement

My econometric analysis shows that aligned districts have 21% higher credit disbursement on average as compared to non-aligned districts over the study period.

However these two categories of districts could differ in terms of other characteristics – besides alignment – that influence credit disbursement. For example the (unobservable) demand for credit may be higher in aligned districts. To address this issue I analyze how credit disbursement changes in 38 districts which experience a change in alignment. For these cases the change in credit disbursement and alignment can be observed for a particular district which overcomes a potential bias in the  results induced by differences in unobservable district-specific characteristics. I find that for this group of districts credit disbursement increases (decreases) by 28% when a district becomes aligned (non-aligned).

Although lead bank of a district does not change change in alignment may occur on account of reasons such as change in convenorship of the state (for example from Union Bank of India to State Bank of India (SBI) in Manipur in 2005 and Allahabad Bank to Bank of India in Jharkhand in 2013) formation of a new state (for example Telangana1 in 2014) or merger of a bank and its associates (for example SBI in 2017). Could the change in alignment of a district be driven by credit disbursement instead of alignment affecting credit as posited here? In Figure 1 I plot the trend of (log of) credit in the quarters leading up to the change in alignment for districts that exhibit a change in alignment (‘treated’) and neighbouring districts that do not (‘control’) experience any change in alignment. The pre-trends appear to be parallel indicating the absence of any unusual change in credit disbursement that could influence the choice of lead and convenor banks.

Impact of negative income shock on long-term savings: Aligned vs. non-aligned districts

Next I observe the impact of a negative income shock on long-term savings. Specifically if aligned districts do have lower bottlenecks to access credit then the ability to borrow easily should ameliorate the impact of an exogenous (external) economic downturn on long-term savings (Eswaran and Kotwal 1990).

I use scanty monsoon (20% below normal) in a district as proxy for a negative income shock and quarterly term deposits2 as a measure of long-term savings for the period 2012-20163 After a scanty monsoon (third quarter of a financial year) aligned districts should have access to credit which should prevent their long-term savings from getting depleted. Consistent with this hypothesis term deposits in non-aligned districts are found to fall by 25% in the quarter after a scanty monsoon but there is no such reduction in term deposits for aligned districts.

Could monsoon rainfall be correlated with unobservable district-level characteristics? If aligned districts are more susceptible to scanty monsoon rainfall then households may be better prepared for consumption smoothening which may bias the above result. A further statistical test of difference in district-wise departure of monsoon rainfall from normal levels rules out this concern.

Conclusion and policy implications

In India’s Lead Bank scheme each district is assigned a public-sector bank as lead bank to improve district-level financial inclusion and each state is assigned a public-sector bank to monitor the lead banks. My study finds that when district- and state-level agents of this scheme belong to the same corporate entity credit disbursement is 21% higher. The results are explained by lower cost of monitoring of the state-level agent which induces higher effort by district-level managers.

Whether greater credit disbursement is good or bad from a social welfare perspective is a separate matter. This requires assessing ‘adverse selection’4 or ‘moral hazard’5 in allocating credit at the district level. To do so one would need data on district-level non-performing assets (NPAs) which to my knowledge is currently unavailable in the public domain.

This limitation notwithstanding the results shed light on how organisational design may influence organisational performance. Complex organisational set-ups operate in the various welfare schemes in India such as distribution of subsidised fertilisers LPG (liquefied petroleum gas) distribution Pradhan Mantri Fasal Bima Yojana (crop insurance scheme) UDAN (regional aviation connectivity scheme) and so on. Implications from theories of organisational economics may help understand and address distortions and ultimately improve public service delivery.

Notes:

1. The state of Andhra Pradesh was bifurcated to form two states – Telangana and the residual region of Andhra Pradesh – in 2014.
2. District-wise monsoon rainfall was available only for 2012-2016.
3. Term deposits are long-term saving instruments which offer high risk-free returns.
4. Adverse selection occurs when buyers and sellers in a market have unequal information and this can distort the usual market process and lead to market failure.
5. Moral hazard is a situation in which one party gets involved in a risky event knowing that it is protected against the risk and the other party will incur the cost. It arises when both the parties have incomplete information about each other.

It’s time to address the root cause of the 1992 stock market scandal – a conflict of interest in RBI’s many roles

The release of a recent Sony LIV series on the Harshad Mehta financial markets scandals that rocked India in the early 1990s has triggered chintanam on an event that perhaps led to some of the biggest reforms in the Indian financial sector. The “securities scam” was in essence a diversion of funds estimated at between Rs 3500 crore and Rs 5000 crore during 1991-92 from the inter- bank government securities (G-Sec) market to stock market manipulators. These diverted funds were used to artificially create a giant rise in stock prices.

The Government of India and the Indian Parliament mounted a capable response to these events in three phases. In the first phase we got the Securities and Exchange Board of India (Sebi) a modern effective regulator for securities market and two new securities infrastructure institutions — the National Stock Exchange and National Securities Depository Limited — which featured novel governance arrangements. In the second phase we got a modern derivatives market enabling risk mitigation and improve price discovery. The third phase saw the convergence of Sebi and the Forward Markets Commission a step towards unifying all organised financial trading in India into one unified set of exchange and regulatory institutions.

Somewhere in this journey we lost focus on the root cause of the 1992 scandal. The heart of the problem in 1992 was the inter-bank G-Sec market. At that point in time the Reserve Bank of India (RBI) through its Public Debt Office Department (PDOD) managed public debt. As the Joint Parliamentary Committee (JPC) on the subject noted the failure of the PDOD to discharge its responsibilities provided scope for irregularities in transactions in government securities. The malfunctioning of PDOD led to the misuse of “Bankers Receipts (BRs)”. The JPC concluded that “the misuse of BRs could be traced to the inefficient functioning of the PDOD and the lack of effective supervision by the RBI”. It further said that “the supervisory mechanism in the RBI was only in form and lacked substance and effectiveness.” Quite a strong indictment!

Having the central bank manage public debt generates a series of conflicts which have negative effects on economic and financial policy. There is a conflict of interest between setting the short-term interest rate (i.e. the task of monetary policy) and selling bonds on behalf of the government. If the central bank tries to be an effective debt manager it would lean towards selling bonds at high prices i.e. keeping interest rates low. This leads to an inflationary bias in monetary policy. Half of this conflict has been solved by legislatively mandating an inflation-targeting RBI. But the other half of the problem namely the RBI as the debt manager of the government remains.

There is a further conflict of interest in bank regulation by an RBI which is also handling public debt. When the RBI tries to do a good job of discharging its responsibility of selling bonds it may veer towards using its regulatory power over banks to force banks to hold more government bonds. Having a pool of captive buyers undermines the growth of a deep liquid market in G-secs with vibrant trading and speculative price discovery. It is too easy for the RBI to sell government bonds to captive buyers in normal times but then in exceptional times (such as 2020) the lack of a diversified pool of voluntary buyers of bonds hampers an expansion of the deficit.

The failures of achieving a liquid and efficient government bond market have hampered the development of the corporate bond market which requires the foundation of a tradeable and hedgeable government bond yield curve. When the central bank controls the market infrastructure for the G-Sec markets as the RBI currently does this creates another conflict where the owner/administrator of these systems is also a participant in the market.

Separating debt management from the RBI is key to addressing these conflicts. In this framework the RBI focuses on monetary policy i.e. on the task of modifying the short-term interest rate so as to stabilise the domestic business cycle. The debt management office works as the “investment banker” for the government selling bonds and engaging in other portfolio management tasks in close coordination with its client the budget division of the Ministry of Finance. Each of these agencies then has a clear focus specialised professional skills are built up (in monetary policy at the central bank and in investment banking at the debt management office) and conflicts of interest are minimised.

This is hardly a novel set of ideas; this intellectual clarity has been present from the early 1990s onwards. The Narasimham I Committee on Financial Sector Reform (1991) clearly identified the conflicts that result when the same institution manages debt and regulates banks. The Working Group on Separation of Debt Management from Monetary Management which submitted its report to the RBI in December 1997 recommended the separation of the two functions and establishment of a company to take over debt management. The RBI Annual Report for 2000-01 recommended the separation of the functions of debt and monetary management in the medium-term and the explicit removal of the debt management function from the RBI. The first significant move on debt market reforms was taken by the present government in 2015 when it brought a draft law to  Parliament but later withdrew it. Mainstream thinking on this question seems to have shifted since then. There is greater clarity at the RBI that it must only pursue the 4 per cent inflation target and divest other responsibilities. For example Professor Amartya Lahiri who used to head the Centre for Advanced Financial Research and Learning recently wrote that redesign of India’s financial architecture requires three reforms of which one is “the RBI needs to be relieved of its public debt management role”. He draws extensively on the recent books by Urjit Patel and Viral Acharya. In the present RBI governor we have a person who has actually handled each of the conflicts discussed above and lived through the difficult task of reconciling them.

This confluence of factors not least of which is the bulging public debt suggests that this is a good time to get back to focusing on and solving this problem.

The writer retired as a secretary to GoI and is now a professor at the National Council of Applied Economic Research. Views are personal.

The urban poor is still finding it difficult to return to work. Targeting social safety nets towards them is necessary as the economy struggles to recover.

While COVID-19 continues its assault on human lives the Indian economy after the devastation in the wake of the lockdown is showing signs of recovering. Data from a variety of sources such as exports and car sales as well as data from NCAER’s Business Expectations Survey point to the onset of a recovery. A new employment survey in the Delhi-NCR region fielded by NCAER adds to this chorus. However it also points to the unevenness of this recovery and the continued vulnerability of specific populations.

NCAER’s National Data Innovation Centre initiated a monthly telephone survey of men and women in the age group of 21-59 in March 2019 to improve the measurement of women’s work. The sample for the survey was drawn to be representative of the 31 districts falling within Delhi-NCR. This includes both urban and rural areas some as far away as Bharatpur and Jind. This survey would have ended in April this year but with the onset of lockdown after a brief hiatus it was extended to measure the impact of the lockdown and its gradual relaxation. Telephone interviews with about 2200 adults who were interviewed monthly between March 2019 and September 2020 paint an interesting picture. While every individual did not respond to each monthly survey contrasting employment in the second quarter of the calendar year (April-June) and third quarter (July-Sept) for 2019 and 2020 for the same population allows for robust year-on-year comparison.

Four observations from this study are noteworthy. First the employment data shows definite signs of a recovery. Comparing employment patterns of individuals in 2019 and 2020 over the April to September period shows that a sharp decline in employment took place in the second quarter due to the lockdown as well as signs of a recovery since the relaxation of the lockdown restrictions. For men the worker-to-population ratio (WPR) in the second quarter fell from 88 per cent in 2019 to 62 per cent in 2020 — a fall of 26 percentage points. In contrast in the third quarter of 2020 the WPR was only 5 percentage points below that in 2019. For women the comparable decline was 16 percentage points in the second quarter regaining most of this lost ground in the third quarter and remaining only 2 percentage points below its 2019 level.

Second self-employment has emerged as a protective force. Individuals engaged in farming and small household businesses have weathered the employment slowdown better than those engaged in salaried employment or those employed as casual labourers. Self-employment in the second quarter of 2020 was down by 12 percentage points for men and nine percentage points for women but by the third quarter it had returned nearly to its 2019 levels for both men and women. Microbusinesses experienced considerable distress in the early phase of the lockdown but they seem to be easing. In contrast wage employment for men declined by 19 percentage points during the lockdown and remains about 7 percentage points lower in the third quarter; for women comparable figures are 10 and two percentage points below the 2019 figures for second and third quarters respectively though women start from a much lower level of participation in wage work.

Third the immediate and lingering impact of the lockdown has been uneven and larger employment declines are observed in urban areas. The year-on-year WPR difference for urban men was 35 percentage points in the second quarter and 10 percentage points in the third quarter. While for rural men the deficit was 15 percentage points in the second quarter down to two percentage points in the third quarter — an almost complete recovery.

Fourth individuals at the bottom of the income pyramid have been affected by job losses far more than the individuals at the top of the income pyramid and this difference is striking in cities. The WPR dropped by a whopping 44 percentage points in the second quarter and remains 13 percentage points below its 2019 levels in the third quarter for urban men who were in the bottom three quintiles of the household asset ownership before the pandemic. In contrast for the top two quintiles the decline was 27 percentage points in the second quarter rising thereafter to remain only 8 percentage points below their 2019 levels in the third quarter.

Even as the threat of the disease persists Delhi-NCR significantly affected by COVID-19 early on is showing signs of an economic recovery. However this recovery is uneven and the most vulnerable urban residents seem to be the last to recover. They were also the ones most affected during the lockdown. As the Delhi-NCR Coronavirus Telephone Surveys (DCVTS) show the urban poor particularly the urban informal sector workers showed the greatest signs of distress during the lockdown and reported greater hunger indebtedness and an inability to pay rent.

The virulence and spread of COVID-19 have justified the early preventive actions taken by the government. As large parts of the country struggle to contain the pandemic the experience of Delhi-NCR holds important lessons. Social safety nets like MGNREGA and PM Kisan have historically been targeted towards rural residents. While a rural bias makes sense under normal conditions the lockdown has disproportionately affected urban workers particularly wage workers who had few assets to begin with. The urban poor is still finding it difficult to return to work. Targeting social safety nets towards them is necessary as the economy struggles to recover. These findings add to the urgency of thinking about an urban employment programme analogous to MGNREGA.

This article first appeared in the print edition on October 202020 under the title “An urban safety net”. Desai is professor University of Maryland and director NCAER National Data Innovation Centre Deshmukh and Pramanik are research associate and deputy director at NCAER NDIC respectively. Views are personal

The new MPC has done well but it would take game-changing reforms for our economy to regain its pre-covid trend growth

The new monetary policy committee (MPC) which submitted its first report last Friday has started well. Faced with an unprecedented crisis a steep gross domestic product (GDP) contraction of 24% in the first quarter of 2020-21 along with inflation rising above the Reserve Bank of India’s (RBI) tolerance band to nearly 7% the MPC chose to navigate a cautious path between stimulating economic recovery and containing inflation. It has maintained all its policy rates including its repo rate (held at 4%) to contain inflation. Meanwhile RBI has taken several measures to ease liquidity lower the yield on the benchmark 10-year government security (G-Sec) and lower the private sector’s cost of borrowing to stimulate economic activity. The MPC and RBI have correctly judged that elevated inflation is a transient phenomenon driven largely by supply-side disruptions during the lockdown. Given the exceptional circumstances they have focused on the steep economic contraction rather than their principal mandate of containing inflation. (I should disclose here that MPC member Shashank Bhide is my colleague at the National Council of Applied Economic Research or NCAER).

Going forward RBI has forecast that the economy’s contraction will moderate in the months ahead with growth turning positive by the fourth quarter (Q4) of 2020-21 but annual GDP will still decline by 9.5% this year compared to 2019-20. Official projections are typically over-optimistic which helps anchor expectations positively. This is understandable. Hence a 9.5% contraction forecast suggests that RBI has internally considered a double-digit contraction. Governor Shaktikanta Das has hinted as much mentioning there are downside risks to this forecast. A 9.5% contraction is nevertheless a bold projection. It provides a compelling official rationale for deploying both monetary and fiscal policy to revive the economy.

At the NCAER we had assessed by 17 May that GDP would contract by 26% (NCAER Quarterly Economic Review: 2020-21) in this fiscal year’s first quarter (Q1). It was later confirmed that the economy indeed contracted by nearly 24%. We are now forecasting that while the economy has been recovering since Q2 output levels are still much lower than in the same period last year. The year-on-year contraction will moderate from -23.9% in Q1 of 2020-21 to -12.7% in Q2 -8.6% in Q3 -6.2% in Q4 resulting in -12.6% for the whole year (NCAER Quarterly Economic Review 2020-21 Q2 25 September 2020). This moderation in the relative contraction from -23.9% in Q1 to only -6.2% in Q4 actually marks a very sharp recovery the so-called “bow string effect” of the steep decline to a very low base in Q1. It accounts for the many references to a V-shaped recovery. But it also implies that the annual real GDP level in 2020-21 will be 12.6% lower than the real ₹146 trillion GDP of 2019-20.

The key question is how the economy will perform from there on. To assume that real GDP will reach its 2019-20 peak of ₹146 trillion even by the end of 2022-23 would imply a swing in the year-on-year growth rate from -12.6% this year to 7% next year a positive growth swing of about 20% followed by a 7% growth again in 2022-23. Nothing like this has ever been seen in India’s growth history. It is more likely that the 2019-20 real GDP level of ₹146 trillion will be reached again only in 2023- 24. But assuming very optimistically that it is reached again by the end of 2022-23 thanks to the very low base affect of 2020-21 and assuming even more heroically that the 7% growth path will be maintained over the long-term beyond 2022-23 the NCAER Q2 review demonstrated that we would catch up with the output level of a counterfactual “no pandemic” trend growth path of 5.8% only by 2039-40. Of course to assume that the Indian economy can quickly recover from a 12.6% contraction and continue to grow at 7% per year for the next 20 years is just a pipe-dream.

The purpose of all this growth arithmetic is to make the simple point that the long-term effects of a major economic shock or economic hysteresis is really very long lasting. So narratives about the Indian economy soon getting back to a high growth path are unwarranted if not irresponsible. To achieve a new normal of even 5-6% trend growth would be quite challenging. Despite a combined fiscal deficit of nearly 14% of GDP factoring in the 12 October stimulus package plus extra-budgetary borrowing plus liquidity infusion of around 9% of GDP such conventional stimulus policies are unlikely to get the economy back to that growth path. That would require deep economic reforms across a wide front no less far reaching than 1991’s game-changing reforms.

Sudipto Mundle is a distinguished fellow at the National Council of Applied Economic Research and was a member of the 14th Finance Commission. These are the author’s personal views