I’ll just pop in the comments I emailed 🙂

Indeed, the issue involves looking at a range of factors in more detail than I did in a cheeky little blog post. However, we have to be careful with how we split all these sorts of issues when trying to figure out our counterfactual – my goal was solely to make the point that such scheme could also increase competition in the face of existing tacit price collusion. Also, I didn’t intend to give the impression it added to ALL consumers welfare – just that consumer surplus overall was higher. Those are very different concepts of course.

With price discrimination for example, we need to ask who’s demand will be more responsive to the implied wedge between prices that comes from the fuel voucher. The supermarket and fuel station are entering into this scheme because it offers them a way to increase profits, and so in terms of this argument we need the people receiving the voucher to be more responsive to price changes – that way ex-post the supermarket and fuel stations will set higher prices. If the voucher is actually working the other way, I find this argument less compelling – and would in turn put more weight on it being some form of “commitment” towards competition.

Furthermore, the “externalities” are irrelevant for thinking about this specific example. All costs and benefits pertaining to the market are captured in the market in this case – any direct externality from fuel consumption shouldn’t be counted as a cost of the market response, but instead as a direct cost of fuel consumption. This isn’t to say the costs aren’t relevant – it is just to say that they are only relevant insofar as there is a separate market failure that should be dealt with, presumably with a tax on fuel. A tax that in NZ is estimated to be about right.

…

Just to clarify my point on the externalities (as I was likely very unclear), we know that the voucher lowers the price for some, and that its existence increases the price for others – we don’t know whether the effective price is lower or higher. In fact, in the absence of the “prisoner’s dilemma” the effective price is likely to be higher given that market demand is inelastic and the firm is increasing profitability by price discriminating. In this case, relative to the Pareto-optimal allocation we would need a smaller fuel tax in the face of these vouchers rather than a larger one.

The data, however, shows clearly and consistently that vehicle ownership is strongly and positively related to income (Figure 4, Litman T., The Future Isn’t What It Used To Be) When you add up the poor people who can’t afford to drive, the young (half of 16-19 y/o don’t even have a full license), old and disabled who can’t drive, and those who choose not to drive, it’s hard to see how vouchers are not a wealth transfer from those that are poor, disabled, can’t or choose not to drive.

Further, your discussion did not take into account the cost externalities of petrol subsidies if they turn out to lead to more driving which includes more crashes, injuries, congestion, poorer health, vehicle expenditures and local and global pollution, among other established costs.

Explanation 2 might explain the initiation of these price discounts, but the continuing program is probably more explanation 3. 

“But I reckon the fun answer would involve the interaction between monopolisticly competitive supermarkets and monopolisticly competitive petrol companies.”

FYI, I do agree that this would be the fun answer 🙂

Industrial economics problems are definitely the most interesting economic problems you can get – even if they may not have the gravity of macroeconomics in some peoples minds.

Very true, if the two firms just competed on price there would be no scope for schemes that rely on such inefficient income transfers – and the only real explanation would be price discrimination (or a behavioural explanation, which I am steering away from).

I was attempting to say that welfare would be higher than it is in the case without vouchers – where you seem to be suggesting that welfare is lower than it would be if the firms competed (if I am interpreting correctly).  I agree 110%, and don’t think the agruments are inconsistent – furthermore, if we had the welfare optimum it is true that the voucher scheme would be unncessary, I also agree there.

If I wanted to strech my argument I could say: why we can’t appeal to an incomplete form of tacit collusion between the firms here – they tacitally collude on price, but then can compete on the type of schemes they put in place.  Or I could simply assume that the offer of a voucher scheme from petrol companies is a mechanism that can be used to breakdown tacit collusion by changing incentives for the supermarket – and its existence then implicitly increases competition.