Two questions about the 2G scandal

The scandal involving corruption in the allocation of 2G licenses left the nation stunned. This column presents estimates of the revenue loss to the government, and says that while the common man was affected indirectly as social spending may have been reduced due to the losses, there was no direct negative impact on the telecom market in terms of lower quality of services or higher prices.

The 2G scandal marked a watershed moment in India’s recent governance history. As the largest of a series of concurrent corruption incidents – for example, the Commonwealth Games scam, the housing loan scam, amongst others – the sheer scale of the money involved stunned the nation. Moreover, the lurid details of leaked tapped conversations, businessmen in cahoots with the media which itself was in cahoots with the government, were both lapped up by and outraged the public. The scandal led to the arrests of the Telecom Minister Mr. A. Raja as well as a number of corporate bosses, presumably cost Mr. Raja’s party subsequent elections in his home state of Tamil Nadu, and sparked an anti-corruption movement drawing support from the hitherto apathetic Indian middle class. The Supreme Court eventually declared the entire process of license allocation flawed and voided the licenses allocated by Mr. Raja.

Now that the dust has settled, it is time to take a closer look at what the actual consequences were of corruption in the allocation of 2G licenses. As I4I editor Ashok Kotwal noted in an email, the scandal has generated much heat but no light. In this column, I attempt to cast light on the consequences suffered by the government and the consumer, answering two key questions: what was the revenue loss to the government and how consumers were affected, if at all.

The 2G scandal

For those who did not eagerly follow the details of the scandal, here is some background on the telecom industry in India and what actually happened starting late 2007. To operate wireless mobile service in India, one needs both a license and radio frequency spectrum, both allocated by the Telecom Ministry. Licenses are region-specific, with the country split into 22 roughly equally sized regions, and an initial license entitles companies to a standard frequency block. In 2007, the Department of Telecommunications (DoT) was considering 575 license applications from 46 companies.

The procedure used to process these applications and whittle down the actual number of licenses awarded to be commensurate with available spectrum was almost comical in its arbitrariness. First, there was no auction and no price discovery: the DoT used prices fixed in 2001 even though it was January 2008 and telecom markets had witnessed unprecedented growth in the interim.

Next, notwithstanding a two year delay in processing applications, an October 1 deadline was suddenly announced on September 24, 2007. Amazingly, this deadline was ex-post reset to September 25 on January 10, 2008! Finally, the first-come-first-serve queue was compromised by announcing – at 2:45pm on January 10 – that the queue would only hold if bank guarantees (of hundreds of crores of rupees!) were provided between 3:30-4:30pm that very day! Eventually, 122 licenses were awarded to 17 companies, many of whom had no previous experience in telecom, and all of who received a substantial discount¹.

Subsequent investigations by the Central Bureau of Investigation (CBI) revealed that the DoT’s procedures weren’t merely whimsical, but rather carefully calculated in order to benefit companies alleged to have bribed Mr. Raja. It took until May 2010, when conversations between Mr. Raja and a corporate lobbyist were revealed by the media, that public pressure forced these investigations to pick up steam. After the Comptroller and Auditor General (CAG) produced its report detailing irregularities in the licensing procedures and calculating the loss to the government in November 2010, Mr. Raja was forced to resign. He was finally arrested in February 2011, over three years after the license allocation. In February 2012, the Supreme Court declared the allocation “illegal” and voided the sale of licenses, ordering the government to create a new procedure to allocate licenses.

What was the loss to the government?

The most widely cited figure of the loss to the government from the sale of licenses and 2G spectrum was the CAG estimate of Rs 1.76 lakh crore (or Rs 1.76 trillion, for those not versed in the Indian system of higher numbers, or Rs 1760000000000, for those who just like seeing large numbers written out). This figure is patently absurd. It is based on the auction of 3G spectrum that was conducted in April 2010. This is akin to pricing a 2008 Maruti 800 the same as a 2010 BMW 510 and justifying this by saying they are both cars. Just as baffling is the fact that the media focused on and kept repeating this figure with barely a shred of protest².

At the other end of the spectrum (!) is the figure derived from the re-allotment of licenses through an auction in November 2012 - Rs. 9,400 crore, almost exactly what they were sold for in 2008. A (somewhat) smaller hue and cry arose when the licenses were sold at prices that were so much lower than the losses cited above, with the Congress and its allies attacking the CAG for its original figure. But it is just as ludicrous to use the re-auction prices as an estimate for what the government lost in 2008. The outlook for telecom markets as well as the Indian economy looked far different in late 2012 than it did in early 2008. Bidding companies would have had far more information about potential revenues and profits given the changes during the period – for example, the 3G auction in April 2010, as well as the global recession³.

Between these extremes, what is a reasonable estimate for the losses suffered by the government? Amazingly, one is buried within the same CAG report that had the Rs 1.76 lakh crore figure. This estimate is based on the dilution of equity by companies subsequent to acquiring licenses. The most sensible estimate comes from the equity sale by Swan Telecom, given that Swan was a complete shell company⁴, with no physical or human capital assets. The value of the company was basically the value of the licenses, its only assets. Since Swan sold 50% equity to Etisalat for approximately Rs 3,600 crore, the value of the 13 licenses it owned was approximately Rs 7,200 crore. Adjusting this figure for the regions in which Swan acquired licenses – some regions are more valuable than others – and extrapolating to the full set of 122 licenses sold, the CAG arrived at a presumptive loss to the government of Rs 57,666 crore⁵.

There are a number of reasons why this figure is a good estimate, although ideally the value of the equity would be discovered on the market rather than through a private sale. First, there is no reason to expect that Swan would not want the highest possible price for its licenses. Second, if dilution of equity by another company – Unitech – is used as a benchmark, the CAG obtains a slightly higher figure but one in the same ballpark: Rs 69,626 crore. Given that Unitech had some assets, unlike Swan, this upward adjustment makes sense. Finally, another benchmark is provided by prices offered by S Tel Ltd, one of the original licensees, in November 2007 prior to the license allocation: using these prices, the CAG arrived at a loss figure of Rs 67,364 crore, again in the same ballpark.

So what does Rs 57,666 crore buy you? While only about a third of the higher figure, it will still feed a lot of mouths. It is roughly equal to the entire amount spent on the Targeted Public Distribution System (TPDS) - India’s massive food distribution program, and over a third higher than the entire expenditure on MNREGA - the country’s flagship workfare scheme, in 2007-2008. Given these comparisons, the uproar over the 2G scandal seems entirely justified.

What was the impact on telecom markets?

The loss to the government meant that the common man was affected at least indirectly as social spending may have been reduced. Did he also suffer directly through lower quality service or higher prices for wireless telecom? The answer here is far more sanguine. Using data from the Telecom Regulatory Authority of India (TRAI) as well as industry associations, I find that the corrupt allocation of licenses in January 2008 had no impact whatsoever on telecom markets (see Sukhtankar 2013 for details).

The empirical analysis this statement relies on is relatively straightforward. The CAG has determined which companies were ineligible and hence should not have been awarded licenses. In addition, the CBI has charge-sheeted top officials from a number of these companies for illegally influencing the allocation process. Given that licenses were region-specific, some regions will have more licenses awarded to these “corrupt” firms than others⁶. I ask whether telecom indicators in these more affected regions were relatively worse after the new license allocation when compared to less affected regions.

Using this “difference-in-differences” approach, I initially find that the number of wireless subscribers as well as firm revenues were actually higher while quality of service indicators were better in the more corrupt regions after the new license allocation. This would suggest that the corrupt allocation had a benign effect on telecom markets! However, this naïve approach does not take into account differences in pre-existing trends in these outcomes across the regions. Once these trends are accounted for, I find absolutely no differences in the number of subscribers, prices per minute, average minutes used, firm revenues, or quality of service.

Many previous analyses of corruption find that it has harmful effects on firm performance and economic activity (Olken and Pande 2012). Why were things different in this instance? I suggest that at least two factors may have played a role in the wireless telecom case. First, although like Swan a number of shell companies or those with no expertise whatsoever in telecom received licenses, these firms then sold large stakes to experienced telecom entities like Telenor of Norway and Etisalat of the UAE. Second, the high degree of competitiveness in the wireless telecom sector meant that these new entrants had to provide services efficiently or suffer failure. After the new allocation of licenses, the number of operators went from an average of 6.6 to an average of 12.2 per region, while indicators like the Herfindahl-Hirschman Index (HHI) also showed massive increases in competitiveness. Under these conditions, the ordinarily deleterious effects of corruption may have been mitigated.

Concluding thoughts

In addition to impacts on government finances and telecom markets, the 2G scam could of course have other, harder-to-measure negative impacts. These could include for example, the breakdown of trust in economic relations, loss of faith in government and public figures, and a decline in market confidence. Silver linings may include the anti-corruption movement created in response to this and other scandals, although what this movement actually accomplishes is anyone’s guess.

Given the state of India’s institutional development, it is certain that other large scams will emerge in the near future. Already, Coal-gate and Chopper-gate have provided grist for the media mills. While the outraged public demands action, policymakers must be careful to watch out against reactionary policies that deter economic activity. For example, defense contracts are on hold in the wake of the chopper scandal (Lakshmi 2013), while construction activity has ground to a halt in Mumbai after officials started cracking down on corrupt industry practices (Bajaj 2012). Striking a balance between transparency measures that help prevent future malfeasance but don’t unduly shackle economic actors will be key to a positive governance environment in India.

Notes:

Full details available in Comptroller and Auditor General (CAG) Report 2010.
The Indian media were not the only culprits; amongst those parroting the high figure (converted to $40 billion) were the New York Times, the BBC, and the Economist.
Telecom Minister Kapil Sibal summarised the situation by suggesting: “it is very dangerous to extrapolate 2010 prices to 2012. The nature of the market in 2008 was different from that of 2010. The state of the market in 2008 and 2010 is different from that in 2012.” (Philip and Aulakh 2012)
A shell company is a company which serves as a vehicle for business transactions without itself having any significant assets or transactions.
Note that this figure also includes losses to the government from allowing companies currently holding licenses for GSM to start CDMA services (or vice versa), or dual technology licenses, as well as from selling additional spectrum to current licensees. It is not clear that these additional licenses/ spectrum should be priced the same as new licenses/ spectrum. The loss to government solely from the sale of new licenses (based on the Swan equity dilution) is Rs 33,230 crore.
Most of the variation in the number of corruptly allocated licenses is related to the amount of available spectrum in January 2008, particularly in areas where the defense forces require additional spectrum to operate.