The regulators and policymakers need to be aware that the short-term gain for the treasury might turn out to be welfare depleting in the long-run
By V Sridhar & Heikki Hämmäinen
Though TRAI gave its recommendations on spectrum pricing way back in August 2018, we can only expect any action by the government during the latter half of the year due to the forthcoming elections. Since spectrum auctions have garnered bountiful collections for the government exchequer, it is worth pitching on a cautionary note, especially when the country is going to take a plunge into 5G and related technologies. Radio spectrum is a critical scarce resource required for operators to deploy mobile broadband, including 5G services. The past auctions, including the last one held in 2016, do not portend well. In the last auction held in 2016, there were no takers at all for the 700 MHz spectrum; only about one-fourth of the spectrum that was put on the block was sold. The total winning price was just about one-tenth of the sum of reserve prices fixed for all bands.
While the government plans for an auction, possibly later this year, including a 3.3-3.6 MHz spectrum as recommended by TRAI for 5G deployment, it is worthwhile to study how spectrum allocation and pricing have been happening around the world. The European Commission recognises efficient allocation and optimal use of spectrum as primary goals. However, a recent study by the GSM Association points out that over the 4G era (2008-2016), the price paid for spectrum increased 3.5-fold in many of the European countries, while average reserve prices increased over five-fold. The report also points out that most of the awards suffered due to high winning prices.
There is one country that is an outlier from the above and that is Finland. Finland is one of the European countries that departed from the rest in allocating radio spectrum for commercial mobile services using a beauty parade method way back in 1999. Though the country shifted to an auction-based method for allocation of spectrum subsequently, the reserve price and winning bid prices have been one of the lowest in Europe. For example, Finland concluded auctions for 3.4-3.6 GHz spectrum in October 2018, with about 390 allotted to 3 operators at an average price of €0.024 per MHz per population. Compare it with the reserve price for 3.4-3.6 GHz spectrum recommended by TRAI of €2.5 per MHz per population (using purchasing power parity conversion) for Mumbai and the lowest of €0.25 for Himachal Pradesh.
There are several lessons to be learnt by the Indian regulator and policymakers from Finland. Firstly, maximising revenue from the sale of spectrum is not the primary objective for the Finnish government. Many researchers have pointed out that higher spectrum prices have resulted in higher consumer prices. In India, though this is not true, due to intense competition on one hand and the low price strategy of the new entrant, results are evident as incumbents are making huge losses.
In Finland, the government consistently sells spectrum for low prices while maintaining healthy competition, to promote technology investment, lower consumer prices, and fast data traffic growth. As a result, Finland has ten times higher data volumes per subscriber compared to other European countries. Though RJio’s greenfield entry strategy with a 4G-only network, and comparatively under-used spectrum, has spiked the data volume consumption to a value very similar to that of Finland, the other incumbent operators lag behind.
Secondly, the release of usable spectrum in anticipation of demand is the need of the hour. Spectrum scarcity results in higher prices and uncertainty for the operators. For example, Finland allocated a substantial 390 MHz in the 3.5 GHz spectrum to three operators. India often puts on block insignificant amounts of spectrum, especially the critical 800, 900 and 1800 MHz ones. As a significant departure, as per TRAI recommendations, it may be possible to put on block about 300 MHz of spectrum in the 3.3-3.6 GHz spectrum. This will provide a substantial contiguous spectrum block for each operator which is essential for high speed 5G networks.
In India, 4G deployment in rural areas is yet to happen. A sub-GHz band, especially 700 MHz, is very much required for the semi-urban and rural rollout of 4G services. The 700 MHz band with its superior propagation characteristics at reasonable prices is required for the operators to roll out 4G services in rural areas of the country. As per the latest TRAI recommendations, even in Himachal Pradesh, the reserve price is about €3 per MHz per population.
In summary, there has to be market-determined “truthful bidding” in auctions to obtain the desired results. It has been well proven by economists and game theorists alike that the simultaneous multiple-round ascending (SMRA) auction mechanism that is employed in India will derive the true value of the spectrum from potential buyers and is one of the methods advocated for allocating scarce resources such as spectrum. Further, the ranking rules, eligibility points assignment, and activity levels as designed in the auction platform are carefully drafted, thus minimising the possibilities of the winner’s curse, collusive bidding and other such imperfect market behaviour. This theoretically strong mechanism works best when left to the market forces to play the game. Hence, TRAI and DoT must work together to set reasonable reserve prices in order to kick off the future auctions. The regulators and policymakers need to be aware that the short-term gain for the treasury might turn out to be welfare depleting in the long-run.
The writers are Professors at IIIT-Bangalore and Aalto University, Finland, respectively