Total Pageviews

Monday, June 25, 2012

In Re: Why Are Government Subsidies Harmful

Arvind Ilamaran, also of GCPP2 - with the added benefit, I think, of GeoEconomics - put up this post on his blog, Indian Liberal @ Wordpress:

From what I remember of our orientation (notoriously unreliable memory here) the GCPP aims to be, or foster, a forum for dialogue. This is me joining in & being all dialogue-ey.


Dear Arvind,

In a word: Disagree : ]
(Sounds cooler in French. Je rejette. Je suis en desaccord. See?)

First off, I'm not sure a subsidy could be classed as a public good. Maybe this is my limited understanding, or just a side effect of trying to read Econ after 3 AM. Bear with me, though? I'll use a classic Public Good as the point of comparison - "Fresh Air". 
(Urban inhabitants get used to putting that in quotes, of course.)

Fresh air is non-excludable. You cannot so design the good that it can be delivered selectively to some but not to others. Fresh air is non-rivalrous. Consumption by one person, even to his/her heart's content (technically, lungs' content) will not significantly affect availability of fresh air for another/others. Of course, this is only true because fresh air (~ oxygen) is available in abundance, so the scarcity criterion (which creates rivalry) does not apply. It is available in abundance, because the production of this resource outstrips its consumption.
(Assume no pollution or climate change. Don't you love economics?)

A government subsidy meets neither of those criteria. Neither a direct subsidy (where government sells at a discount) nor an indirect subsidy (where the government transfers resources) is non-excludable - i.e. it CAN be so designed as to be available only to some members of the population, and not others.
 A subsidy CAN be universal (e.g. on fuel prices), or it CAN be exclusive. At ration shops, proof of economic standing (or lack thereof) can be asked for before subsidised prices are offered. With direct transfers this is even easier, because an entry barrier (BPL status) is built in to the system. Theoretically, we could have price discrimination in fuel sales also - display BPL card to get subsidised petrol, else pay international procurement price equivalent.
(Assume scrupulous honesty, i.e. no corruption, forgery etc. Don't you really love economics?)

Subsidies reflect a direct expenditure of government wealth. Either the government incurs - as you say - a loss while selling, or it gives away money (or equivalent good/services) outright. Unless the government's money is as abundant as fresh air, there IS a scarcity criterion in play. Since the government has limited streams of revenue, there is no abundance of government money. And yes, the government can always print more currency, but I assume we're talking about money in real (value equivalent) terms, not my-wheelbarrow-is-worth-more-than-the-bushel-of-notes-in-it terms. And as we've seen, where a resource is limited, rivalry is inevitable.
(Assume no inflation? Economics doesn't love you back, not that much anyway. Sorry.)

Second, the statement that subsidies necessarily mean making losses carries an implicit assumption that there is an equilibrium market price representing true value of the commodity, e.g. as determined in perfect competition. Market prices may result from various other factors. Given a monopolistic (or, more realistically, oligopolistic) market, government CAN acquire below market price (as sought to be set by the seller/sellers) with no real loss (only a reduction from notional super-normal towards normal profit) as long as its price is greater than / equal to average cost to the producer(s). Even in competitive markets, the government does set a Minimum Support Price for various crops. Farmers then bargain a higher price for themselves in the market - & sell to the government stocks not sold in the market at this reserve price, i.e. less than market price. Government could, theoretically, sell those crops at a profit while still providing a subsidy, by selling at a price point between the minimum support & market prices.
(Assume bargaining power for farmers. Economics loves farmers. The Indian government does, too.)

For that matter, would you call waiver of VAT a subsidy? Prices would dip by 12.5% in most such cases, because the "market price" included a tax component now discounted. In that case, too, government would face only a notional loss of potential tax revenue, not an actual loss since there would be no government expenditure. In fact, where the government levies indirect taxes, it can procure at pre-tax prices & sell with a partial tax waiver, thus making a profit per unit sold without gypping its citizens while STILL offering a subsidy (read as discount).

These are only theoretical scenarios. In practice, subsidies do involve expenditure of government resources - let's take your figure, between 2 to 3% of GDP. Given that most subsidies (though by no means all of them) can be seen as part of Poverty Alleviation Programmes, one would expect this money to come from tax receipts. For instance, one would assume that expenditure on the Right To Education is paid for from revenues raised by levying Education Cess on transactions or commodities. One would also believe that the logical source for such revenues, at a central level at least, is personal / corporate income tax, since the relation between indirect taxation & price levels are direct & proportionate, but between a universal progressive income tax & price levels less so. Also, there is the option of increasing tax enforcement / realisation without raising tax rates. By some estimates, write-offs in the corporate tax segment exceed Rs. 25 lakh crore! (See P Sainath, - table.)

[Also taking off from Sainath, an option would be a degree of fiscal discipline & reallotment of resources within the government - something actually mandated by the government in recent "austerity" measures:

Not that I agree with this caricaturing of the BPL calculation - that's a policy decision with its own set of implications - but his point on probity & public trust is entirely valid. Especially when you read both articles together & realise the true levels of both, poverty & available resource base. Under the circumstances, it's ironic to call India unfriendly or unfavourable to institutional investment, domestic or foreign. Takes a special kind of budget to do favours for corporates & STILL discourage them from investing, no?

Generally speaking, I do agree with both your characterisation of an ungoverned subsidy-inflation vicious cycle, & Amogh's comment on the role of Quantitative Easing in instances of stagflation - which we may be close to facing. A government faced with difficulties increasing money supply in the short run given already low interest rates may indeed turn to Quantitative Easing, buying stocks / bonds etc. from banks or in the market. The biggest problem with this policy is not so much its impact on long term interest as that there is no guarantee that banks will make this increase in reserves available to the public. So the government will have incurred debts, driven down interest (as you say, affecting investors in those assets) but produced a relatively low impact on market money supply. Inflation, but no stimulus / growth. Now we're really screwed, no?
(NEVER assume implementation of economic policy objectives by private / non-state elements of the financial system. For that matter, even assuming State apparatus does implement it can be risky. It's the money system, not a mathematical problem.)

All of which is to say, the answer to your question is simple. Why are government subsidies harmful? Because they are applied inappropriately. They are applied according to only one criterion - political gain (read as votes at nearest election). They *should* be applied with respect to the government's economic capacity (resources available), their sustainability, their short-term impacts on economic performance, their long-term impact on structural incentives etc. Subsidies can be a powerful equity promotion tool, protecting the poorest from the ravaging effects of the inflation/unemployment paradox inherent in economic growth. Of course, if we accept Mankiw's proposition that moves to promote equity are automatic disincentives to efficiency (I'll need a lot more convincing that it's an automatic consequence given either economic or psychological impacts) then another relevant criterion becomes the efficiency loss involved - an acceptable level must be computed. Equally importantly, the assessment metric for the impacts of such measures must also be adjusted for such effects.

Government subsidies are not harmful. They are costly. The question of whether that cost is worth the gains they provide is a policy decision. The question of what costs & what gains are being taken into account in making this decision is for the field of policy analysis   & critique.


  1. I haven't done my readings well and its only past 3 in the afternoon but tell me before I make another attempt to have a go at this post - when you say "...Mankiw's proposition that moves to promote equity are automatic disincentives to efficiency..." do you mean equality or fairness? No sarcasm intended :)

    1. Mankiw talks about Equity in the specific sense of GDP per capita, or redistribution of income. One of his Principles (in the first chapter, Principle 7, on government intervention) is that government will either seek to increase the size of the pie (efficiency / net GDP) or change how the pie is divided (equity/per capita GDP).

      He also holds that intervening to promote one will necessarily drive down the other: if you try to enlarge the pie, some people will get much bigger slices. If you try to slice it more equally, the pie itself shrinks. Which I'm not convinced about yet!