Why is it regressive? High earners pay more for ad-funded sites than low earners. That's not regressive. It's not progressive either, strictly speaking, but it's better than subscription fees which are regressive because everyone pays the same absolute price regardless of their disposable income.

I'm not certain this is true. And targeted advertising certainly makes it less true, not more. Advertisers are not excluding low-income users from advertising, instead they're targeting more products at them that those users are more likely to buy.

Untargeted ads for higher-cost luxury products might make the Internet cheaper for low-income users, but that's not what is happening in the ad industry right now. Ads exist to get you to spend money, and they are just as targeted at poor people as they are at rich people.

I wouldn't be that surprised if the effect is the opposite, since poor people have less access to research resources, comparison shops, and trials that would allow them to combat the psychological effects of advertising. They're also generally under a lot more stress and time-pressure when they shop than rich people are, which is likely to make them even more vulnerable to manipulation during a difficult purchase decision.

I'm pretty sure it is, because nothing changes the fact that you can ultimately only spend what you earn (even considering credit). A share of that spending goes to advertising.

If someone can spend 10 or 20 times as much as another person after basic food and shelter then that difference trumps all other factors by a very wide margin.

I don't think it matters much, but just for the record: I don't believe rich people comparison shop as much as lower income earners. They buy what they fancy and they throw away what they don't like. I also don't think rich people are much harder to manipulate (but I'm not sure about that). They think much less before tapping that buy button. I know that much.

I'm not sure I follow? Google ads don't cost a percentage of the final product price. How much I spend on advertising might be entirely unrelated to the per-item cost of my product -- and how profitable my company is might not have anything to do with how luxury my company is, it might just be down to market penetration and my profit margins. Plenty of companies make enormous amounts of money targeting poor people.

Does an ad for a five-star restaurant on Google cost more than an ad for Taco Bell? And it's not like rich people are being shown a larger quantity of ads on a website than poor people.

If that is true, then per person contributions to ad funded services would be roughly propotional to personal spending. Someone spending 10 times as much as another person would also pay 10 times as much for using Google search or Youtube.

If these services were subscription based then both would pay the same price in absolute terms, which is very regressive in comparison.

But just to make this clear. I don't claim for a second that my extremely crude calculation is anywhere near correct. What I'm saying is merely that subscription funding is very regressive compared to ad funding.

I don't know what the exact extent of that difference is and I cannot break it down to the level of pricing specific Google ads.

> If these services were subscription based then both would pay the same price in absolute terms, which is very regressive in comparison.

This is the part where you kind of lose me, unless a rich person is also watching 10x as many videos on Youtube. It seems like you're saying a fixed percentage of a person's spending is going to the websites they visit, but why would that be the case?

You will see the same number of ads on a Youtube video regardless of whether you're rich or poor. And the cost of each of those ads -- the amount of money that gets paid out by the business -- is based on the competition for the ad slot, not the price of the product. I wouldn't take it as a given that products like Taco Bell and Pepsi spend less on online advertising than a luxury watch brand. If anything, I would expect products in crowded consumer markets (ie lower-cost, non-specialized, mass-market goods and services) to have more competitive ad slots that cost more money to target.

So I understand that rich people spend more money, I agree with you on that. But I don't see how you're connecting that fact to the idea of more money from those rich people going to the websites who are displaying ads. I don't see the thread of logic that says that a product costing $500 per unit is contributing more ad money to a website than a product that costs $5 per unit.

Very roughly yes. Some percentage of a typical company's revenues is spent on ads, and revenues from each customer are obviously proportional to that customer's spending. It's the same thing (leaving aside sales taxes).

Companies try to maximise the effectiveness of their advertising campaigns. The effectiveness depends on how many people actually go ahead and buy the product relative to how much the ads cost. If running ads on Youtube is less effective, then ads prices on Youtube would have to fall and Youtube would earn less.

Let's say only extremely poor people were using Youtube. None of them would ever buy a high-end smartphone. How much would high-end smartphone makers pay to Youtube for the honor of running ads there? The answer is zero.

Now let's say there are two groups of Youtube users. One group never buys a high-end smartphone. The other group buys one every year. Now it makes sense for smartphone makers to fund Youtube through their ads, but only the group actually buying smartphones pays for it. So the rich group effectively subsidises the poor group's Youtube usage.

I have chosen an extreme and unrealistic example to explain the principle. In reality, there will be a mix of products. The cheapest ones will be bought by almost everybody in roughly the same quantities, and some luxury goods are never advertised on Youtube at all. But the relationship between per person spending and that person's contribution to ad funding for the sites they visit still roughly holds.

This is what I think. I'm not an economist though. There are certainly many open questions as to how strong this redistribution effect is and what the effect of ad targeting is. But the claim that there is no such redistribution effect at all seems extremely implausible to me.