I'm an expert on how non-economists think. That's because every year I try to teach 300 non-economists to think like economists.
Non-economists think in terms of income effects. Economists think in terms of substitution effects. That's a stereotype that isn't 100% true, but it still contains a lot of truth.
Here's how non-economists think:
1. Why does the demand curve for apples slope down? "Because if the price of apples rises, people can't afford as many apples."
2. Why does the Aggregate Demand curve slope down? "Same as apples. Because if the price of everything rises, people can't afford as much of everything."
3. Why is inflation bad? "Same as AD. Because if the price of everything rises, people can't afford as much stuff. Duh."
4. Why does the labour supply curve slope up? "It doesn't. It slopes down. If real wages fall people have to work longer so they can still afford to buy stuff."
5. What is the effect of an increase in interest rates on demand? "If interest rates go up, people can't afford to borrow as much and spend as much, so demand falls." (Except for a few much older students, who argue "If interest rates go up, we get more income, and can afford to buy more.")
What's wrong with thinking like this? OK, they missed out the substitution effects, and so ignored the fact that when the price of apples rises relative to other goods, people substitute away from apples and into other goods. But they still got it half right, didn't they?
No. They got it wrong.
Let's start with apples. For every apple bought there's an apple sold. If the price rises by $1, the buyer is $1 poorer, but the seller is $1 richer. Sure, there are many more buyers of apples than there are sellers of apples. But so what? If 1,000 buyers buy 1 apple each from one seller, there are 1,000 buyers each $1 poorer and one seller $1,000 richer. Even if we assume apples are a normal good ("normal" means you buy more as your income rises), there is no way of knowing what the total income effect will be. Each buyer is a tiny bit better off, and will want to consume a tiny amount more apples. But the seller is a lot worse off, and will want to consume a lot fewer apples. Total desired consumption could go up or down, depending on whether the buyers or sellers of apples have the higher income elasticity of demand for apples. It all depends on whether the buyers or sellers have a greater propensity to spend any extra income on apples.
(There are two ways to treat the apple seller's own demand for apples: we can subtract it from the supply curve; or we can add it to the demand curve. You get the same results either way, if you are careful. But it's a lot easier if you add it to the demand curve, because that way the supply curve doesn't depend on the seller's own preferences for consuming apples.)
The thing that caused the price of apples to increase might have an income effect. A late frost, which reduces total apple production, makes total real income lower. If the frost destroys 1,000 apples, worth $1 each, the apple seller is $1,000 poorer at the existing price of apples. But the price increase itself has zero effect on total real income. It merely redistributes income from buyers to sellers.
In aggregate, price changes do not have income effects. They have distribution effects. We can talk about the income effects of price changes at the individual level, once we know whether the individual is a buyer or a seller. At the aggregate level, looking at the market demand curve for apples, we can only talk about distribution effects. And substitution effects, of course.
If all apples were imported, it would be different. The Canadian demand curve for apples would have an income effect. We have excluded all sellers. But it would be different again if some apples were exported. An increase in the price of apples would now increase Canadian real incomes, and the income effect would tend to increase the quantity of apples demanded by Canadians.
The way non-economists think about the demand curve for apples doesn't just leave out the substitution effect. It's just plain wrong. It is simply not true to say that if the price of apples rises we can't afford to consume as many apples. That "we" ignores the seller of apples. If the "we" includes both buyers and sellers, then collectively we can afford to consume exactly the same amount of apples as before.
What works for apples works for everything else. For every apple bought there's an apple sold. For every unit of GDP bought there's a unit of GDP sold. For every $1 borrowed there's $1 lent. For every hour of labour sold there's an hour of labour bought. At the aggregate level, there are no income effects of changes in prices, interest rates, wages, etc. There are only distribution effects. And substitution effects.
We know that substitution effects are negative. That's why demand curves slope down and supply curves slope up. We need a lot more information to tell which way distribution effects will go. They might be negative, or positive, or zero. One of my own teachers once said "Distribution effects are the last refuge of a charlatan". I wouldn't go quite that far, because sometimes you do have some reason for believing they will go in a particular direction. But the onus is on you to make that case. And if you can't make that case, you had better concentrate on the substitution effects.