The First Law of Demand
In Russell Roberts' excellent essay on why incentives matter, he repeats an important economic assumption that I think is often misunderstood: people are self-interested.
There's various equivilent ways to make this point:
- People respond to incentives
- People are self-interested
- People are purposeful
- People are rational
- Demand curves slope downwards
But note that people are selfish, and people only care about money are not on the list.
A decent micro course will make it clear that the first law of demand doesn't imply that people always respond to price changes, just that there is a possible price change that will create a change in behaviour. This point isn't controversial by-and-large, and pretty much everyone acknowledges that prices and quantity demanded are inversely related. But there are extreme conditions which people (typically anti-theorists) believe undermine this claim, and I think they can be put into three categories:
- Is there a little whore in all of us? When Kerry Packer was bidding for the voting rights to screen ICC Cricket, he famously said "There's a little whore in all of us gentlemen, name your price". I do think that this is a valid assumption to make about human behaviour, and the typical criticism isn't that it's innacurate, it's that it is regrettable. Non-economists might accept that in practice people do respond to incentives, but such incentives illicit socially harmful outcomes. I don't see it this way, because the assumption is simply saying that people will be at the table and since this will always expand the menu of choices, it's a Pareto Gain. If someone inherits their mothers house, they might claim that it has infinite value to them. The self-interest assumption says that there is a price at which they'd sell it. Why? Because the consumer has many competing preferences, and in a world of scarcity we make trade offs. Hence selling the house - at the right price - might make enough money to pay for the kids to go to university. Suddenly it's not a choice between mum's old house vs. not mum's old house; it's the house vs. an education. Regardless of whether it's sold or not (quite possibly it's value is so very high there isn't an amount of alternative goods that can bid it away), surely it's selfish not to consider selling? If there's the slightest shred of altruisim within your own preferences, there must be a possible price that would get you to sell. If it's a choice between selling Grandma or a cure for cancer, who's the selfish person? The one who refuses to sell Grandma, or the one who sells her and invests the money into finding a cure for cancer? All we're saying is that people will come to the table: nothing is off limits, we're all open to negotiation. And therefore this is welfare-enhancing, by expaning our menu of choice.
- Luxury Goods A common response is that if prices signal quality, a luxury brand
would fear that a reduction in price would signal a reduction in
quality, and therefore create a fall in quantity demanded. This seems intuitively plausible, but does it undermine the first law of demand? No, because it confuses a change in demand, with a change in quantity demanded. If a company drastically alters it's reputation, it's created a different product. The first law of demand (as represented on a demand curve) applies to the relationship between price and quantity demanded, for a given product. Therefore any other events (any non-price events) are exogenous, and represent a shift in the curve. Even if demand for Skodas has risen, and prices for Skodas has risen, both of these stem from a shift in the demand curve, and not a movement along it. It's possible that the quality of a Skoda has remained constant throughout (but it's worth considering whether this is likely, and if not why not) and the rise in demand is purely due to a price hike and it's corresponsing quality signal. But this is an abstract point, and rests on an assumption that price is used as an accurate indication of quality. If the consumer knows the quality of the product, there's no reason why a rise in price would lead to a rise in quantity demanded. For any given Skoda, if the price falls you're more likely to wish to buy one. For Skoda cars as a whole, an increase in price might alter the type of product it is, under certain conditions of asymmetric information. But this simply means that tastes have changed, and therefore the curve has shifted.
- Giffen Goods It's easy to bat away charges of "Giffen goods" by appealing to (non-existent) evidence, but the theoretical point is straightforward. A "Giffen good" is, by definition, a good where the income effect dominates the substitution effect. In other words it constitutes so much of your shopping basket that price changes have a large effect on your real income. So if it's highly "inferior" (i.e. demand falls as income rises) a fall in price can induce a rise in quantity demanded. But again this violates the cetetris paribus condition, since changes in real income are an exogenous factor. If price changes occur simulatenously with a change in real income, the demand curve will shift. And if it shifts, the self-interest assumption hasn't been violated.
So all three of these theoretical objections to the law of demand fail to hold, and I think there are two reasons why students/professors remain confused:
- A misunderstanding of the nature of theory A theoretical premise is not refutable by evidence - it can only be refuted by better theory. Therefore if we define a normal good as one where demand rises if income rises, and then label coffee a normal good, evidence (hypothetical or otherwise) that a rise in income leads to a fall in demand for coffee is irrelevent. In that case, coffee isn't a normal good. It doesn't mean that normal goods don't exist. Consider the three primary colours of red, blue and yellow. In real life we never see these three colours on their own, since we're always viewing some combination of them all. Evidence of a green object doesn't mean that blue and yellow don't exist, it just means they're not always observable and fixed.
- Confusion between prices and revenues A firm isn't interested in charging as high a price as possible - revenues are what matters. A simple monopolist's cost structure will show that even if he had enough market power to triple his price, this would lead to an increase in costs and probably reduce profits. Therefore it's wrong to apply the laws of demand to firm behaviour, because the laws of demand focus on how consumers respond to prices. However firms don't care about prices, only revenue (and how revenue and cost correspond to generate profit). In fact, it's likely that a monopolist would increase profit by lowering prices (depending on the elasticity of the demand curves).













Recent Comments