Friday, December 19, 2014

Youth education and unemployment

India Together carried this piece. Reproduced below with references

In a widely reported speech, the Prime Minister recently claimed that the 21st century can belong to India as it has three assets that no country has: ‘democracy’, ‘demand’ and ‘demographic dividend’. ‘Demographic dividend’ has become a much bandied phrase in recent years. 

The allusion is to India’s current demographic state where the ratio of the working age population to the dependent population (the children and the elderly) is near its peak, a situation that will last another 20 to 30 years. Having a greater proportion of working age people in the population can mean faster per capita economic growth, but only if their full potential is realised. A look at the state of employment in the country shows that India is at a far cry from deriving any advantage from its demographic dividend.

Data on employment in India is provided by three different agencies - the National Sample Survey Organization (NSSO), the Labour Bureau and the decadal Census. The three agree broadly on the worker and student populations. It is in identifying the unemployed from among the non-working population that they come out with widely varying estimates.

There are several reasons for this. Among them is the difficulty in determining if women engaged in housework, particularly in rural India, are doing so under duress, being unable to find a job. Another is in deciding if youth, nominally enrolled in educational institutions but still looking for a job, are part of the labour force. Problems also arise in deciding if workers who work in casual jobs for short durations are to be considered as working or unemployed.

The census with its count of those not working (even part time) but “seeking or available for work” shows up a far higher level of unemployment than both the NSSO and Labour Bureau sample surveys of contiguous years. In the following discussion, the unemployment numbers are from Census 2011 made available in Sept 2014.

In 2011, over 56 million people aged 15 and above who were not working were seeking or available for work, putting the unemployment rate at 10.7%. This was however not the full extent of job seekers. Just short of a quarter of the work force consisted of “marginal workers” – workers who were able to find work for less than 6 months. Nearly half of the marginal workers were also “seeking or available for work” taking the count of job seekers to nearly 21% of the labour force.

Youth and unemployment

Further insights into unemployment may be gained by looking at the new entrants to the workforce. The transition from education to work happens between the ages of 15 and 29. As NSSO data shows, around 90% of children below 15 are enrolled in school while above the age of 30, the numbers attending educational institutions are miniscule.

The accompanying chart shows unemployment rates derived from the 2001 and 2011 census for different age groups with a focus on new entrants to the workforce. Unemployment rates across all age groups have increased between 2001 and 2011, a period of high growth of the Indian economy! Jobs have clearly not kept pace with the increase in job seekers.

The age – unemployment relationship is also striking. Unemployment is highest in the 15-19 age group and decreases progressively for higher age groups. For youth aged 15-29 as a whole, the unemployment rate is 21% (counting only those seeking or available for jobs among the non-working) compared to 5% for people aged 30 and above. Young men (15-29) account for nearly 85 % of total unemployment among men in 2011! (Young women account for over 61% of total unemployment among women) The new entrants to the workforce bear the brunt of unemployment.

Education and unemployment

The other dimension of youth unemployment that needs to be explored is its connection to education.

A significant trend of recent years is the increased proportion of boys and girls opting for higher schooling as seen in the accompanying chart showing NSSO 2012 data. The trend in favour of education also extends to young adults between 20-24 years who would typically be enrolled in a college diploma or degree course. Over 25% of young men in this age group were studying in 2012! The increased percentage of youth attending educational institutions in 2012 (compared to 2005) was possible because of a lower fraction going for work.

A significant trend of recent years is the increased proportion of boys and girls opting for higher schooling as seen in the accompanying chart showing NSSO 2012 data. The trend in favour of education also extends to young adults between 20-24 years who would typically be enrolled in a college diploma or degree course. Over 25% of young men in this age group were studying in 2012! The increased percentage of youth attending educational institutions in 2012 (compared to 2005) was possible because of a lower fraction going for work.

If enrolment in education is increasing, what is happening to the composition of the new entrants to the labour force? The accompanying chart shows the educational composition of the youth (15-29) labour force and the youth unemployed in 2005 and 2012 using NSSO survey data. (While NSSO surveys underestimate unemployment compared to the census, this is unlikely to affect the trends seen in the data.) 

The comparison of the situation in 2005 with that in 2012 outlines several trends.

One is that the young entrant to the labour force is increasingly more educated. The second is that it is not the illiterate who are most likely to be unemployed. The relatively better educated youth – those with secondary and higher qualification – have a far greater share in total unemployment than in the labour force. In other words, unemployment is higher in more educated sections of youth. Finally, the share in unemployment of the most educated section, the youth with diplomas and degrees, is actually increasing over time.

What makes it harder for youth with greater education to find jobs? Educated youth naturally have higher expectations from the job and clearly there are not enough jobs that can meet these expectations. The less educated youth, most likely from poorer families, are more likely to settle for just any job be it casual labour or poor quality self employment. The high unemployment among the educated youth in particular reflects the lack of growth of jobs in organized manufacturing and services sectors.

In summary what is abundantly clear is that India faces a severe youth unemployment problem with the relatively better educated sections worse off than the rest. The problem has actually worsened in the decade between 2001 and 2011, with the economy unable to expand jobs for educated youth fast enough despite high GDP growth. With trends showing youth opting for longer years of education, the future entrants to the labour force are going to be relatively more educated and face greater employment challenges.

Against this backdrop, there is absolutely no reason for optimism that India will benefit from its demographic dividend any time in the near future. Instead of empty talk, the government needs to present a concrete roadmap of the steps it is taking to promote employment intensive economic growth.

Data Sources:

“Employment and Unemployment situation in India”, NSS report 554, NSSO, Jan 2014
“Report on third annual employment and unemployment survey (2012-13)”, Labour Bureau, Sept 2013
“Employment and Unemployment situation in India”, NSS report 551, NSSO, Sept 2006
Data on workers (tables B1 and B13) from Census 2011 and Census 2001 available at 

Other articles on this subject:

“Unemployment in an era of jobless growth”, N Chandra Mohan, ISID working paper 159, Jan 2014
“Youth employment and unemployment: An Indian Perspective”, A Mitra & S Verick, ILO Asia Pacific Working Paper Series, Mar 2013

Friday, November 14, 2014

Economic Notes: The imbalance between employment and growth

Close to half of employed Indians work in agriculture. The ‘value add’ in agriculture – the amount by which the value of output exceeds the value of intermediate inputs – is an indication of the health of the agricultural economy. In 2012, this was Rs 63,000 per agricultural worker, less than a fourth of the average figure for non-agricultural workers [1]. At about Rs 170 per day, this was barely higher than the minimum daily wage of unskilled agricultural labour in some states. It must be kept in mind that ‘value add’ is only an average figure and that different categories of agricultural workers - large farmers employing labour, small farmers tilling their own land and wage labourers - will have differing claims on the value added.

For the millions trapped in agriculture the possibilities of escaping poverty depend on the rapid growth of agriculture and the availability of jobs in more productive sectors outside. There are major impediments to the former, chief among them being the fragmentation of land into very small holdings, making the mobilization of capital for development extremely difficult. The choice of the latter is contingent upon better more paying jobs being available outside agriculture for the ‘unskilled’ farm worker.

The existential crisis faced by the Indian peasant is not unique. All across the world, the presently industrialized nations, had at some period in their history, a greater part of their workforce engaged in agriculture under conditions of low productivity in comparison with industry. Industrialization progressed with the help of labour migrating from agriculture. This migration in turn opened up the path for growth in agriculture through productivity increases, eventually narrowing the gap in productivity between agriculture and industry. The rapid change in these economies from agriculture accounting for the lions’ share of employment and gross domestic product (GDP) to industry taking its place has been termed a ‘structural transformation or shift’.[2]

Structural shifts from agriculture to industry happened not only in the industrialized countries of the west, but also in Asia – in Japan, South Korea and Taiwan. In most countries after the shift, the share of manufacturing in employment increased to over 30%. In the industrial powerhouses Britain and Germany, this share reached 45% and 40% respectively, at its height.  Since then, there has been a decline in the share of industrial employment in these countries. But it must be noted that second structural shift in employment, this time from industry to services happened only after industry had attained a share of about 50% in the GDP.[3][4]

Data Sources: GDP data [5] ; Employment data [6]

Closer home, China too, it appears, is also on its way to a structural transformation with manufacturing accounting for about 33% of the GDP (industry for 46%).[7] India remains way behind in terms of the development of industry. Manufacturing had a share in GDP of 14.4% in 2011-12 and industry as a whole had a share of 26.7%. The above chart draws attention to the basic structural imbalances in the Indian economy – the great discrepancy between share of contribution to GDP and share of employment in different sectors of the economy. (The share of other non-manufacturing industries except for construction - such as electricity, gas and water distribution - has been clubbed with mining in the above chart)

Has a structural shift begun in India?

Claims have been made recently that a structural shift in employment out of agriculture, though greatly delayed, may finally be in progress in India too. The claim is based on the sharp 10% drop in the share of agriculture in total employment between 2004-05 and 2011-12, and the fact that the absolute numbers employed in agriculture have declined for the first time since 2004. [8]  The chart below shows the relative share of employment of different sectors of the economy over the years. Employment in non manufacturing industries (other than construction) has been omitted as it is insignificant.

Data Sources: Authors calculation using employment numbers from Mehrotra et al (2014)

The chart taken at face value seems to show that Indian peasants are moving out of agriculture at an accelerated pace since 2004. The increase in the share of employment has been mainly in construction sector and to a lesser extent in services. Employment share of manufacturing has increased only marginally. A closer examination of disaggregated NSSO data for agriculture throws more light on the shift out of agriculture.

Women are leaving the workforce

The chart below shows the decrease in recent years in the labour force participation rates (LFPR) - the number of people in a category who look for employment out of every 1000 - for rural and urban Indians aged 15 and above. (The same trends have continued into 2012-13 as per labour ministry surveys.)

Data Sources: NSSO Surveys

One reason for the decreasing LFPR is the increasing enrolment of youth between 15 and 24 in educational institutions. This affects both men and women and more or less accounts completely for the LPFR decrease among men. That still leaves the issue of women leaving the labour force.
Over 80% of women leaving the labour force belong to rural India. The most plausible explanation for rural women leaving agriculture and retreating to household duties runs along these lines. Most of these women are from the poorest rural households. These are the households of marginal farmers and casual labourers from which the men have migrated out of the village for casual jobs in construction. Rural women traditionally work together with their men folk either as unpaid family workers or as casual labour and have stopped working as they can no longer accompany their men folk.[9]

The falling LFPR, predominantly affecting rural India and thereby agriculture means that the fall in the employment share of agriculture is not entirely due to workers shifting to other sectors. A significant part is due to rural girls and women withdrawing from agriculture. Agricultural labour is not being ‘pulled’ to jobs in expanding manufacturing in industrial centres; rather it appears that it is being forced to take up seasonal casual work to avoid further deterioration in standard of living. The households of these migrant workers are still in the villages. For all these reasons, it is difficult to conclude that a structural transformation similar to that seen in industrialized countries is under way.

A jobless growth

Going by the historical experience of other countries, manufacturing jobs are the key to facilitating an employment shift out of agriculture. Is manufacturing job growth stagnant because of poor growth?

The following table shows the average annual growth of jobs and value added (same as the contribution to GDP) in the different sectors of the economy in the most recent period for which both sets of data are available. Data for mining also includes all other non-manufacturing industries except for construction.

From 2004-05 to 2011-12, India’s real non-agricultural GDP grew by an average 9.4% per annum, but employment grew only at 3.5%. Services and Manufacturing were the fastest growing sectors at 10.1% and 8.9% respectively. However they added jobs only at 2.5% and 1.5% respectively. Clearly, GDP growth encompassing both manufacturing and services did not translate into employment growth. The answer to why growth in manufacturing is practically jobless growth must be sought in the structure of Indian manufacturing and its markets and is beyond the scope of the present article.
What is amply clear is that growth of the GDP variety is no guarantee of employment growth. A large part of the poverty ridden agricultural workforce waits the creation of factory jobs to transition out of agriculture. Any government that claims to represent the peoples’ best interests must aggressively focus on creating productive jobs in industry.


Binswanger-Mkhize (2013): “The Stunted Structural Transformation of the Indian Economy”, Hans P Binswanger-Mkhize, EPW, June 29, 2013
Kannan et al (2012): “Counting and Profiling the Missing Labour Force”, Kannan K P and G Raveendran, EPW, February 11, 2012
Mehrotra et al (2014): “Explaining Employment Trends in the Indian Economy: 1993-94 to 2011-12”, Mehrotra S, J Parida, S Sinha and A Gandhi, EPW, August 9, 2014
Rodrik (2013): “The perils of premature deindustrialization”, Rodrik D, Project Syndicate,
RUPE (2014): “A middle class India?, Aspects of India’s Economy, No 58, Sept 2014,


[1] Agricultural ‘value add’ is obtained by dividing the contribution to GDP (factor cost at current prices) of agriculture by the number of people employed in it. The ‘value add’ of non-agricultural workers as a whole was 284,000 in 2011-12.
[2] See Binswanger-Mkhize (2013) for a description of the general characteristics of structural transformations in the industrialized countries
[3] A discussion on the growth in employment in manufacturing during the structural shift in the industrialized countries can be found in Rodrik (2013) and in Rodrik’s weblog
[4] See also chapter 4 of RUPE (2014) for a discussion on the structural transformation thesis and its applicability to India
[5] GDP data is from the Central Statistics Office press release, MOSPI website
[6] Employment data is based on estimates using unit level NSSO data by Mehrotra et al (2014)
[7] See note 2 for numbers on China
[8] Mehrotra et al (2014) write: “Post 2004-05, for the first time in India’s post-Independence economic history, in the five-year period, 2004-05 to 2009-10, as many as 23.7 million of India’s agricultural workforce abandoned agriculture..... In fact, non-agricultural employment grew by 25 million.... Between 2009-10 and 2011-12, non-agricultural employment increased sharply – a 27 million increase in absolute terms, while at the same time the numbers in agriculture fell by 13 million in a matter of two years. This is a historically unprecedented development in India’s economic history.
This structural shift .....  is precisely the kind of progressive structural change in employment that should accompany a structural change in output between the primary, secondary and tertiary sectors in any developing economy.
[9] This is the conclusion drawn in Kannan et al (2012) based on a detailed analysis of data between 2004-05 and 2009-10. The same trends appear to have have continued through 2011-12 and very likely through 2012-13 as well.

Friday, May 9, 2014

Counting the Law

How many central statutes are there in all? Bibek Debroy estimates in this ET piece the number to be between 2500 and 3000. The oldest are the 'Bengal Districts Act' and 'Bengal Indigo Contracts Act' both dating to 1836! There are 140 laws that go back to the 19th century and another 200 that predate independence. No one has a count of state level statues. That means that there is no compendium of all the laws applicable at the state level.

So how do the law enforcement agencies keep track of this huge number? How many statues are actually enforced, even occasionally?

Monday, April 28, 2014

How political parties can accept money without disclosing the source and spend without limit on the elections

This piece was written in April 2009. Everything written here still holds good

Fighting elections in India has become extremely expensive. According to an Election Commission (EC) representative, while an estimated Rs. 4500 crores was spent for the 2004 Lok Sabha elections, this time around, the expenditure will cross Rs. 10,000 crores. About 20% of this constitutes government expense; the remaining will be spent by political parties and their candidates. Though a huge sum, even this is in all probability a conservative estimate.

The high election expenditure is not a reflection of the basic cost of the campaign but rather an indication of the permissive environment that has evolved over time where there is no limit for the amount of money that political parties can collect and spend on elections. Political parties are neither accountable nor transparent about their finances. The state machinery is unwilling or unable to curb illegal expenditure by political parties allowing the elections to become an outlet for huge quantities of black money. Such an environment encourages political parties to outdo each other in spending in the quest for political advantage. Politicians across the country have chartered as many as 60 helicopters costing between Rs. 75,000 and 2 lakhs per hour for use over a month (Hindustan Times, April 21, 2009). Profligate spending is no longer frowned upon. But this was not always so.

The law on election expenses

The election related laws framed in 1950’s had the objective that money should not be allowed to influence the outcome of elections. The Representation of People Act, 1951 required every candidate to keep an account of all election expenditure incurred or authorized by him. The expenditure was to be kept within prescribed limits and subject to inspection by EC representatives. Violating the limit was deemed a corrupt practice and punishable by disqualification from contesting elections for a period up to 6 years.

The law did not set limits on the expense of political parties and there was a reason for this. The Constitution of India did not even recognize the ‘Political Party’ as a formal entity at that time; political parties had no defined role to play, either in the elections or in the formation of government. The law to limit expenses therefore only dealt with the expenditure by the candidate; this was considered sufficient to curb overall expenses on the elections. In time, political parties started exploiting this loophole.

Did expense incurred by the sponsoring political party or by the friends and supporters of a candidate towards his election in excess of prescribed limits constitute a corrupt practice? This issue came up before the Supreme Court in the case of Kanwarlal Gupta vs. Amar Nath Chawla (1974). Using the occasion to expound on the objectives of the ceiling on spending by candidates, the Supreme Court observed:

“It should be open to any individual or to any political party, however small, to be able to contest an election on a footing of equality with any other individual or political party, however rich and well financed it may be, and no individual or political party should be able to secure an advantage over others by reason of its superior financial strength.”

“The other objective of limiting expenditure” the Supreme Court added, “is to eliminate, as far as possible, the influence of big money in electoral process. If there were no limit on expenditure, political parties would go all out  for collecting contributions and obviously the largest contributions would be from the  rich and the affluent who constitute but a fraction of the electorate….. The small man's chance is the essence of Indian democracy and that would be stultified if large contributions from rich and affluent individuals or groups are not divorced from the electoral process.”

After outlining these objectives, the Supreme Court argued that “if a candidate were to be subject to the limitation of the ceiling, but the political party sponsoring him or his friends and supporters were to be free to spend as much as they like in connection with   his election, the object of imposing the ceiling would be completely frustrated and the beneficial provision enacted in the interest of purity and genuineness of the democratic process would be wholly emasculated.” The thus Supreme Court answered the original question in the affirmative.

This judgment met with the negative reaction of the government of the day. The Representation of People Act, 1951 was amended by the Congress government in 1975 and an infamous “explanation” added to the election expense provision. The explanation ran thus:

“Notwithstanding any judgment, order or decision of  any  Court  to  the  contrary,  any expenditure incurred or authorized in connection with the election of a  candidate  by  a  political party  or  by  any  other  association  or  body of persons or by any  individual  (other  than  the candidate  or  his  election  agent)  shall  not be deemed to be, and shall not ever be deemed to  have been,  expenditure  in connection with the election incurred or authorized by the candidate or by his election agent….”

Not only the sponsoring political party but also friends, relatives and supporters of a candidate were freed from any limits on spending for their candidate. The objectives of the original provisions in law limiting election expenses of candidates had been completely frustrated!

Financial transparency of political parties
The Income Tax Act was amended in 1979 requiring political parties to file income tax returns every year. The tax law allowed a political party to claim full exemption from income tax for a variety of sources of income including voluntary contributions received from any person provided the party maintained books of accounts, recorded details of all voluntary contributions above Rs 10,000 and got their books audited. However, most parties including the Congress and the BJP did not comply with the law and file returns. This fact only came to public notice after a Public Interest Litigation filed by ‘Common Cause’ came up before the Supreme Court in 1995. The Supreme Court ordered the Government to investigate and prosecute the erring political parties, but it is not clear if anything came of this. The Law Commission of India in its 170th report, “Reform of the Election Laws”, in 1999, made this scathing comment on the issue:

“While a small income-tax payer who fails to file his return is prosecuted and penalized, the political parties which are in receipt of huge funds which they spend on elections and other occasions are not being touched. The parties too do not appear to have realized that if they themselves do not follow the law, not only it sets a bad example to others, they will not have the face to tell others to abide by law.”

The Commission followed a comprehensive analysis of the problem of election expenses with the recommendation that political parties must be required by law to keep accounts, have them audited and publish them for the general public and strong penalties should follow including de-recognition of the party by the EC for non compliance. The National Commission to Review the Working of the Constitution also expressed similar views in 2002.

The election expense laws were finally amended in 2003, but not on the lines suggested by the Law Commission. The Election and Other Related Laws (Amendment) Act, 2003 was passed by the NDA Government with the support of almost all parties including the Congress. While the infamous 1975 “explanation” to the election expense clause was finally deleted, this change made no difference any longer. The election laws and rules were amended to allow political parties to “accept any amount of contribution voluntarily offered to it by any person or company”. Only contributions to parties over Rs 20,000 (earlier Rs 10,000) were to be recorded and reported. The punishment for not submitting returns was that income tax exemptions could not be claimed! The income tax laws were also amended to give 100% tax exemptions to companies and individuals for contributions to political parties. The NDA government claimed that these changes would bring about more political accountability. But many questions remained unanswered. Would companies, for example – solely in business for profit - contribute a part of their profit to party funds without any expectation of return favors when/if the party came to power?

The idea of limiting election expenses had been buried once and for all by codifying in the Representation of People Act, 1951 the right of political parties to accept (and consequently spend) any amount. Enacting strict penalties to ensure transparency and accountability in finances of political parties had been given the go by, even though it was clear that the laws would not be respected otherwise.

The Election Commission suggested to the UPA government in 2004 that “political parties must be required to publish their accounts (at least abridged version) annually for information and scrutiny of the general public and all concerned, for which purpose the maintenance of such accounts and their auditing to ensure their accuracy is a pre-requisite” Their plea fell on the deaf ears. The UPA was as comfortable with the status quo as the NDA.

Who will fund the parties?

The escalating cost of elections puts pressure on parties to mobilize funds whichever way they can. Most political parties do not collect money for party activities and elections by building a broad membership and collecting regular dues. The funds collected using legal channels from companies and individuals are only a fraction of what they ‘need’, given the possibilities of unlimited expenditure. During the 2004 elections, for example, all the political parties put together showed expenditure less than Rs. 230 crores according to figures made public by the EC. Sources of unaccounted wealth need to be tapped; this is probably what makes parties shy of making their ‘real’ books open to the public. Seen against this context, the recent action of the Bahujan Samaj Party (BSP) in Delhi in nominating extremely wealthy businessmen is understandable. The President of its Delhi unit admits that as the party does not fund its candidate’s election expenses, it expects them to have deep pockets if they should have a fair chance to win. The BSP candidates for four of the seven Delhi constituencies have declared assets of 622, 155, 14.5 and 19 crore rupees respectively!

There has been a concerted campaign for some time past – initiated by certain NGO’s and now enthusiastically adopted by corporate India and the media - to reform the middle class elector. The hope that is held out is that if more people vote and exercise their choice with discrimination, things will improve. The question that begs an answer is who will persuade the political parties to reform?  

24th April 2009

Thursday, April 24, 2014

How India subsidizes cereal exports

Interesting article  by Ashok Gulati in the Economic Times where he points to the increasing cereal production in the country linked to higher production in states like MP and Chattisgarh, the easing of export curbs,and the increasing exports of cereals (mainly rice, wheat, corn). Gulati maintains that the exports are subsidized by the government as it subsidizes the inputs that are needed for cereal production - water, electricity and fertilizers.

Gulati would have the government target the subsidy while freeing up exports completely. But are cereal exports from India at all in public interest? Are there other ways in which the state can intervene to promote the right balance between cereal production and the production of other crops?

Monday, February 17, 2014

AAP's Case for Lower Electricity Tariffs in Delhi

My article with this title appears in the Feb 22, 2014 issue of the Economic & Political Weekly as a commentary. It is reproduced below.

The Aam Aadmi Party (AAP) announced a 50% subsidy for electricity consumers using under 400 units a month within a few days of forming its government in Delhi. While the major political parties criticized the move, the state governments of Haryana and Maharashtra quickly followed suit announcing their own subsidies.

There is a strong public perception in Delhi that bills are inflated and that electricity distribution companies (discoms) have managed this by installing fast running meters and obtaining tariff increases by dubious means. The stiff resistance of the discoms to providing information under the RTI Act or to submit to a CAG audit served to strengthen this perception.The former Delhi Government also compromised itself in the public eye by stonewalling a CAG audit.

The AAP’s election promises spoke to these public concerns and its government has taken steps to fulfil them, including pushing for a CAG audit.AAP’s expectation is that the audit results will provide grounds for reducing the tariff. The subsidy has therefore been announced only for three months and its impact is less than 1% on the Delhi budget.

This commentary takes a critical look at the case made out by AAP for an audit of the discoms and lower tariffs after providing a brief introduction to the tariff fixing process. For more details on the subject of tariff determination as well as the disaggregated components of tariff of Delhi discoms, the reader is referred to Kasturi (2013).

Electricity distribution in Delhi, with the exception of the New Delhi Municipal Council area and the cantonment, is split between the Tata Power Delhi Distribution Ltd. (TPDDL) belonging to the Tata group, and BSES Rajdhani Power Ltd. (BRPL) and BSES Yamuna Power Ltd. (BYPL), both belonging to the Reliance ADAG group.

The tariff process requires the discoms to submit a multi-year plan to the Delhi Electricity Regulatory Commission (DERC) stating their annual revenue requirement that includes the cost of power, operation and maintenance expenses, depreciation, and a guaranteed return on capital employed. After review, the DERC approves the requirement for each utility and fixes a tariff structure that will enable all the utilities to meet their revenue requirements.

The actual performance of a discom in any year may differ from plan because of factors beyond its control. The discom gets an opportunity through a ‘true up’ petition to the DERC to ask for recognition of variation in expenditure and revenue from plan. The DERC, after validating the discom’s claims, issues a ‘true up’ order that spells out the revenue requirement and the revenue actually realized from sales that year, as recognized by the commission. A gap between the two provides the case for a revision of tariff.

‘True up’s’ happen after considerable delay. Discoms in Delhi submit their ‘true up’ petitions for any year, only a year later. The DERC then takes several months to arrive at its order. The last true up orders for Delhi discoms (as of writing in Feb 2014) were passed by DERC in July 2013 and square up the accounts of 2011-12.

For better clarity, the revenue requirement of the discom may be represented by its average cost per unit of power provided to the consumer. This cost is composed of the cost of the power procured, the cost associated with distribution loses (due to theft or technical loss) and the cost of distribution (including the profits guaranteed to the discom). The latter two, clubbed together is termed ‘distribution overhead’ in this commentary. The revenues realized by the discom from electricity sale may be represented, similarly, by an average billing rate.

Table 1 illustrates the average operational costs and billing rates for the three Delhi discoms for 2011-12. The discoms used the gap between average billing rate and operational cost to press for higher tariffs. We will proceed to examine in detail each component of cost and the revenue along with AAP’s contentions about them.

Table 1: Average operational costs and billing rates of Delhi discoms for 2011-12

Power purchase cost A
Distribution overhead B
Total cost C = A + B
Billing rate

All costs and rates are in Rs/unit; Power purchase cost is based on total number of units available at the discom network periphery. Other costs and rates are based on total number of units billed to customers.
Source: Respective 2011-12 ‘true up’ orders

Energy purchase and sales

The bulk of Delhi’s power requirement is met through long term purchase agreements with generation plants in the central and state sectors. Discoms enter the short term power market to meet peak loads and sell surplus power during lean periods. Table 2 shows the power purchase costs for TPDDL and BRPL.

Table 2: Power purchase cost of TPDDL and BRPL


Long term purchase

Short term purchase

Transmission costs and losses





Surplus sales

Net for customers

Cost escalation due to short term trades





Long term purchase

Short term purchase

Transmission costs and losses





Surplus sales

Net for customers

Cost escalation due to short term trades




Rates are in Rs/unit
Source: DERC ‘true up’ orders

Up till 2008-09, surplus power was sold at high rates and more than made up for the higher cost of short term purchases. In 2009-10 and 2010-11, this situation was reversed. Discoms bought larger amounts of short term power at high rates and sold surplus power at lower rates leading to an overall increase in the cost of power of 15-16%.

Based on the high rates obtained for surplus power in 2008-09 and projecting a greater surplus for 2010-11, the DERC in May 2010 was on the point of issuing an order reducing tariffs. The discoms complained against this to the Delhi Government which promptly stopped the DERC from making the order. The High Court, hearing a challenge to this government intervention in the DERC’s work, ruled a year later directing the government to desist from such interventions in future and asking the DERC to proceed with a new tariff order.

The DERC, by now under a new chairman, was presented with the 2009-10 accounts by the discoms which included huge overruns in power purchase costs. It accepted these cost overruns without questioning and increased tariffs through its tariff orders of August 2011.The above events are at the centre of AAP’s assertion that the tariff increases since August 2011 are unjustified.

A year later, in July 2012, the DERC acknowledged irregularities in power trades of the discoms in the ‘true up’ orders of 2010-11 accounts, stating that:

 “a competitive bidding process was not followed and power has been contracted on the basis of request for offers from traders (through verbal communication). As such, there were no supporting documents to validate the selection of the contracts against all the offers that may have been received” (BRPL) and

“Contracts for short term power purchase were finalized very early in the year and for most of them, competitive process was not followed… appears that that no efforts were made to sell through exchange or bilateral sources where better yields are expected.”(TPDDL)

After making these observations, the commission was content to rest the matter with the advice to the discoms that “there were scope for better management of the process of short term power purchase and sale of the surplus power so as to significantly promote the interests of the consumers.”

Another year on, things had still not been set right with power trading as is visible in DERC’s stricture to BYPL in Aug 2013:

“Contracts for sale of short-term power were finalized early in the year and no competitive process seems to have been followed. It is noted that substantial quantum of short-term power was sold through Exchange and UI…..”

These were not the only disquieting practices in power trades. Power purchases and sales of the discoms were often mediated by power trading companies including those belonging to the Tata & Reliance ADAG groups and the ultimate seller / buyer was not revealed.

Clearly the discoms at a minimum were not putting in their best efforts to manage power costs. This would not be surprising, given that the cost of power is a pass through to the customer. The AAP alleges worse - that the discoms converted what should have been a revenue surplus from sale of surplus power into losses in their books. Its election manifesto contains a promise to make the buying and selling of electricity by the discoms transparent and to take steps to bring them under the RTI Act.

Distribution overheads and average billing rates

Distribution overhead has two components as explained earlier – the cost associated with distribution losses and the distribution costs. Table 3 shows the distribution losses and the distribution overheads over the years for the Delhi discoms and for BEST, a public sector Mumbai discom. All four are of comparable size having between 1.1 and 1.7 million customers.

Table 3: Distribution losses and distribution overheads of discoms



Overheads are in Rs/unit and based on the total number of billed units
Source: DERC true up orders for Delhi discoms; MERC for BEST

While the three Delhi discoms had comparable overheads in 2007-08 and 2008-09, BYPL started showing much higher overheads than TDPL from 2009-10 and BRPL the same from 2010-11. These two companies failed to progressively cut down distribution losses, a major reason for their high overheads. Their losses were much higher even in comparison to BEST, a public sector company operating in Mumbai city.

At the time of the privatization of electricity distribution in Delhi, the major benefit held out was that distribution losses would be dramatically cut down. The disincentives for not meeting the distribution loss targets do not seem to have worked for these two discoms.

So what is the reason for BRPL and BYPL failure to cut down their distribution losses?

The AAP alleges that the Reliance ADAG group discoms have been manipulating accounts showing lower billing (and revenue collections) in their books; consequently the calculated distribution losses turn out higher than what they actually are. A higher distribution loss pushes up the cost of power available for distribution to the consumer while lower collection pushes down the average billing rate. A larger gap between the two (see Table 1) would allow a discom to make a case for a larger tariff hike.

In support of its allegations, the AAP has drawn attention to the sudden increase in distribution losses in a number of circles of BRPL in 2010-11 reversing the steadily decreasing trend of several years. It has also pointed to the extremely damaging findings of the DERC during the validation of the 2010-11 billing of BRPL and BYPL.

The commission found numerous instances of customers with zero consumption or zero rate billing and many customer categories with average billing rates lower than the prescribed tariff in a sampling of the billing database of these two discoms. Summing up the validation exercise in connection with BRPL the commission noted:

The information provided by the Petitioner during the entire validation session was inconsistent and changed many times. …. the Petitioner could not provide clarifications to the satisfaction of the Commission.

The findings of the DERC from the validation exercise relating to the 2010-11 true up throw up the question of how a regulator should deal with a discom if data presented for a true up petition is unsupported by evidence or does not stand up to scrutiny. DERC penalised BRPL by an estimated Rs 52 crores in 2010-11, just over 1% of its revenue requirement for that year. The AAP would have the regulator take stringent action, including prosecution for cheating, in case of serious discrepancies in the accounts presented to the commission.

A visit to the DERC website, the repository of all data relating to tariff fixation, is utterly disappointing. There are no metrics or time series data available to help the public judge the performance of discoms. The latest true up petitions of the discoms of 2013 that would contain all the data submitted to the DERC, are not available. The most recent minutes of the meetings of the ‘State Advisory Committee’, a body mandated under the Electricity Act to advise the commission on matters including “protection of consumer interest”, date to Oct 2011.

The discoms have provided enough grounds through their questionable accounting and power trading practices to justify an independent audit of their accounts. In Delhi, they are monopoly providers of an essential public good using infrastructure set up in the course of decades by the state. Leaving aside legal merits, it is entirely reasonable that they come under the ambit of CAG audits and the RTI Act.

The issue however goes beyond an audit to uncover wrongdoings of the past. How are consumer’s interests to be safeguarded in future? If consumer charges are decided on the basis of cost and a regulated profit, surely there should be transparency into the energy and financial accounting of the discoms, at least to the extent that these have a bearing on the tariff. Indeed the Electricity Act provides adequate powers to the State regulatory commissions to make regulations on the information required to be disclosed by discoms. It remains for the government of the day to find ways to persuade the DERC to bring in this transparency.

An earlier article, also published in EPW, on electricity pricing in Delhi is available here