5 ms·
Very little is certain in economics. Possibly the best thing for our economic future would be unemployment in excess of 63%, high enough for the two programmers
by wnewman 15y ago
Very little is certain in economics. Possibly the best thing for our economic future would be unemployment in excess of 63%, high enough for the two programmers in Atlanta who currently carry in their brains the complementary halves of an all-important tech insight to meet in an unemployment line, put their heads together, and revolutionize our economy! But that's not the way to bet.
In general ordinary limited assets like spectrum space, or agricultural land, or airport landing slots, or electricity, or food can be used better by some actors than others. If your economy doesn't allow them to be traded freely, it's an extremely good bet that the economy will throw away important efficiencies, because whoever currently has the right to control the asset doesn't have good incentives to pass it on to someone with a really good use for it. Thus e.g. human food being fed to pigs, or water being used to grow rice in a desert, or various ways that rent-controlled properties get underutilized. You seem to be assuming that the allocation of spectrum will be like that (that now that they're "spread more widely" that pattern of control is fixed in stone). If so, the Indian economy can expect to lose a lot of the potential value of the spectrum because people who in 2012 have really effective ways of using it can't get to it, and it instead it gets used in whatever way is technically feasible for the organization which was ushered to victory in the race through the permit Raj ca. 2008.
This isn't just some obscure academic possibility. Consider various historical uses of spectrum in the US. E.g. various religious organizations had valuable spectrum they couldn't sell freely, or even use freely, but they were allowed to use it to broadcast sermons and the like, so they did. I dunno what the analogous lockup of assets in India will look like in detail, but in broad outline the gross waste could easily be similar.
Conversely, if you allow spectrum to be traded freely, that disadvantage goes away, but then the "spread more widely" pattern that you consider an advantage, or for that matter anything else you like about the distribution of handouts, tends to go away too. The runout effect tends to be what another poster already pointed out, that the sham auction effectively gave the assets away for arbitrarily low prices to selected customers, and now those selected customers will be able to collect a fortune by exchanging their control of those assets for something closer to the assets' market price.
- barrkel 15y agoI think you're putting words in my mouth: words like "handouts" and "fixed in stone". But with respect to "spread more widely" that you quote twice, for common resources that is explicitly the goal of democratic government; most efficient use of the resource is not (it might be in anarchy, or oligopoly, or some other system not based on one person, one vote). Let's not forget that auctions won't necessarily predict the market price; in fact, they're unlikely to, because in this kind of case they rely too much on prediction of future value, and the structure of such licenses will be that they won't trade freely on a market, so normal pricing mechanisms won't work well. There are lots of inherent rigidities: hardware dependencies limit the pace of change. If a government is to get the windfall, it would probably be better (IMO) to tax profits at a later stage. My perspective is coloured by some of the worse effects auctions of public telco assets had in Ireland; we ended up with telcos deep in debt, owned by foreign private equity (effectively purchased with the company's own debt), starved of revenue to invest and almost completely focused on servicing the debt. I remember at the time the alternative method being used in Sweden[1] and Finland, focused more on so-called beauty contests than auctions. [1] http://www.sciencedirect.com/science/article/pii/S0308596105000558 http://www.sciencedirect.com/science/article/pii/S0308596105...