September 18, 2006

Mi Negra: A good idea that's being oversold

Manuel Rosales has made two headline proposals during his campaign:
A-A pledge to pay the unemployed a minimum wage.

B-"Mi Negra" - a plan to issue lower and middle class people a debit card to distribute 30% of the nation's oil revenues directly into people's pockets, in payments that would range between Bs.600,000 and Bs.1 million per family per month.
One problem is that the relationship between the two pledges is murky - I for one have not seen a detailed proposal - so it's not so clear if the money for A is supposed to come from the same 30% of oil revenues that will fund B.

I have argued elsewhere that Pledge A is just absurd. I don't think I need to spend any more space on it here.

Mi Negra, however, is not absurd. In fact, conceptually, it's very appealing: it would break the petrostate model at the root, by transcending the oil-money-for-political-support quid pro quo that is the crux of petro-populism. Conceptually, I like it a lot.

My problem is that the proposal lacks detail, so I'm reduced to doing back of the envelope calculations to try to say something about its viability. What I find is that the only way Rosales can deliver payouts of Bs.600,000 - Bs.1 million per family is to leave millions of poor families out of the program. But any attempt to make Mi Negra highly selective would bring very serious problems of its own.

Tallying up Mi Negra
Granted, everything that follows depends, needless to say, on the assumptions you make. But a quick-and-dirty calculation of Mi Negra's viability might go like this:

Assume Venezuelan oil exports fetch $38/barrel, we export 2.2 million barrels per day, and production costs hover around $8/barrel. Then, 30% of export earnings works out to:

$30 per barrel x 2.2 million export barrels x 365 days x 0.3 = $7.2 billion/year

Now, consider that nearly everyone in Venezuela is either poor or near-poor. The per-capita figure, then, is:

$7.2 billion / 25 million = $289 per person per year.

Assuming a standard 5-person household, the per family per month pay out would be:

$289 x 5 people x 2,150 Bs:$ / 12 months = Bs.258,000 per family per month.

That's less than half the Bs.600,000 Rosales is touting as his bottom-of-the-range estimate. Even if you chose to limit payments to just half of Venezuela's families, you'd still come up short.

Even when oil prices are very high - like now - the numbers don't really seem to work.

In 2006, for instance, VenEconomy estimates that PDVSA's total fiscal contribution will come to Bs.80 trillion, a hefty $37.5 billion.

30% of $37.5 billion comes to $11.2 billion. So at a time of sky-high oil prices, the amount to be handed out via Mi Negra comes to:

$11.2 bn / 5 million families / 12 months x 2,150 Bs:$ = Bs.400,000 per household per month.

Which is still one third short of Rosales's pledge of Bs.600,000 minimum.

The Perils of Selectivity
Now, the obvious solution here is to introduce stricter selectivity, giving money out to the poorer half of the population only. Parts of Rosales' website suggest this is the plan. The sum would then be:

$11.2 bn / 2.5 million families / 12 months x 2,150 Bs:$ = Bs.800,000 per family per month.

Selectivity is an intuitively appealing solution, but one with serious practical problems. One is that, these days, the Venezuelan middle class is so small that excluding half the population means shutting millions of poor families out of the system.

A bigger problem is that the moment you introduce selectivity, you introduce perverse incentives.

Families with incomes just below the (by nature arbitrary) threshold would pocket the Mi Negra money and leapfrog ahead of families with incomes just above the threshold. The leapfrogged families will not take this sitting down.

They will have a strong (perverse) incentive to either hide their income from the authorities - easy, given the scale of the informal economy - or to work less so they qualify for Mi Negra. As more and more families do this, the average payout from the system would tend to fall.

What's more, it's easy to foresee that whatever government agency is charged with deciding who is above and who is below the Mi Negra threshold would have a strong tendency to become corrupt. The more selective the program is, the more bureaucrats you need to run it. And bureaucrats given discretion over decisions worth hundreds of dollars to millions of people are very well positioned to pocket a cut.

But isn't Rosales selling Mi Negra as the simple, non-bureaucratic alternative to oil rent redistribution?

In the worst case scenario, Mi Negra could come to be used as a new mechanism for Petrostate populism, where loyal Rosalistas get the subsidy and his opponents do not. And, well, cabra tira pa'l monte - given Rosales' roots in the AD petro-populist system, I don't think it's crazy to think Mi Negra could degenerate along those lines.

So selectivity is costly on a number of levels, and could end up subverting the whole point of the system.

I think Mi Negra is a laudable idea, but the devil is definitely in the detail. A less selective program would be far simpler and cleaner, but it would offer substantially less money to each recipient family. To make the program more generous you have to make it more selective, but then you create layers of perverse incentives and run the risk of creating a bureaucratic monster. Hard choices have to be made here - there is no magic bullet.