Free public transport, equality, means testing brain, and Jago Dodson's very bad research conclusions that he should retract

Recent events have made the free public transport debate extra salient. Trump attacked Iran, war broke out, oil/gas/fuel/petrol prices are up, and in Australia, temporary free public transport has been announced in two states (Victoria and Tasmania) but ruled out in New South Wales.

The free public transport debate has long fascinated me, because many participants don’t take the side one might expect. Specifically, people you would describe as public transport “experts”, “advocates” or “supporters” oppose it, often very strongly - not only claiming that it isn’t a priority, but that it is actively bad, on two main grounds:

• “Free PT would mean less money for services”: Sacrificing fare revenue at the same time as increasing demand (due to price no longer being a barrier) means less money and more strain on existing services, so worse services and passenger experience.

• “Free PT is regressive and bad for equality because it primarily benefits wealthier people”: Public transport coverage is better in cities and in the wealthier, more established areas of cities, and used typically by white collar commuters. On this basis, free PT represents a handout to the wealthy who already enjoy relatively good public transport.

The problem is that these claims are wrong, bogus, and built on a foundation of bad research that generates wrong conclusions from conceptually confused thinking.

Worse, they stem from the same bad conceptual thinking that has poisoned welfare state discourse in Australia and worldwide for decades now, which we might call “means testing brain”.

Means testing brain

The conceptual problems stem from not making apples-to-apples comparisons, not considering both sides of “the ledger”, and muddying absolute quantities vs percentage quantities.

So let’s detour from public transport fares to the welfare state and consider the situation with old age pensions.

Many countries, Australia included, means-test benefits like old age pensions, meaning people with more than a certain “means” (ie income/wealth) are excluded from receiving it.

This is superficially appealing from an egalitarian and efficiency perspective: Don’t give government benefits/spend government money on those who don’t need them, ie those who have the “means” to do fine without it.

It seems like that gives us less government spending/lower taxes (the government saves a bunch of money not handing out money it doesn’t need to) and better “targeting” (more of the money goes to people who need it, and less to those who don’t).

But all this is wrong. The key missing insight is that the same total tax is being collected, just in a different form. Not getting paid a pension is the same as being paid a pension but being taxed that same amount. In both cases, I receive zero dollars.

So comparing a means-tested pension to a non-means-tested pension isn’t an apples-to-apples comparison. To do that we need to look at the full system, or both sides of the ledger. Taxes and benefits.

Another way to think about this clearer is to decompose means-testing into two separate components: A universal pension, and the means test “tax”.

Expressed in basic mathematical notation, the total cost of each system (ignoring administration costs for the moment) is:

Means-tested pension system = universal pension - means test “tax”

Universal pension system = universal pension - regular tax

Again, for a given government expenditure, the amount of tax is the same.

In other words, the universal pension vs means-tested pension debate is simply a difference of:

• Distributional effects (winners/losers of 1 vs the other)

• Administration costs

It doesn’t reduce overall tax - it just shifts it and disguises it.

Means testing is just a weirdly-shaped tax with high admin costs

So if the distributional effects of means-testing were good, and the administration costs were low, it would be fine.

But sadly the opposite is true.

There are numerous problems distributionally:

• Non-payment of a benefit is a flat tax of an absolute+limited amount, rather than a percentage like a regular tax would be. Consider a billionaire like Gina Rinehart. Percentage-wise, not receiving the pension is a very low effective “tax” for her, but a very high “tax” for somebody with only just enough assets to lose out. With a regular progressive (or even flat) tax, Gina would pay a higher percentage, not a lower one.

• Regular taxes are raised from lots of different people, and lots of different sources, meaning lots of different people, retirees and non-retirees alike. But non-payment of a benefit is a “tax” only on retirees, and further, only on a subset of those retirees. Because the “tax” is less “wide”, it has to be “higher”, much higher. The effective tax rates, percentage wise, of real-world means testing are ludicrously high, much higher than regular taxes ever are or would be, and often over 100%!

• Administration costs are higher (means testing everybody is more expensive than simply handing out to everybody) which means more waste and less money going to recipients

• Administration is a barrier to participation, often to those worst off and least able to navigate bureaucracy.

If you’re still unconvinced by how much of a stinker means testing is, I highly recommend these two great reads:

• Universal Benefits Cost Less Than Means-Tested Benefits by Matt Bruenig atPeople's Policy Project

• Means testing is a dog of a tax and it will destroy the welfare state by David Sligar

Public transport fare brain is just another variant of means testing brain

Switching gears back to the public transport fare debate, we see all the same conceptual problems.

• “Slashing fares would mean higher taxes” ignores that fares are already effectively a tax. Swapping out fare revenue for regular taxation wouldn’t be higher taxes, because saving on fares and administration costs would be an effective tax cut.

• “Free fares would benefit the wealthy” ignores that the wealthy would pay much more if the same fare revenue was raised via conventional taxation instead. This is doubly true for the very wealthy who tend to get around in limos, helicopters etc rather than trains and buses.

• “Fare revenue means more money for services” is wrong, because, per above, the same revenue can be gathered with regular taxation instead with lower administration costs. Free fares actually means billions saved in ticketing and enforcement costs that can be used to fund better services instead.

Fares also suffer from some even worse conceptual problems than pension means testing, distributionally speaking:

• At least with pension means testing, the worst off are at least theoretically eligible for the pension and excluded from the non-payment. With PT fares, even very poor people must pay, and even if they are eligible for concession rates, those concession fares can comprise a substantial percentage of their income.

• Whole vs parts fallacy. Just because PT is more prevalent in wealthier areas, and oriented around commutes to white collar CBD jobs, doesn’t mean that everybody who takes public transport is wealthy. Obviously, plenty of poor people take public transport, in particular those too poor to afford a car.

Free fares lower inequality and ease poverty, and Jago Dodson is flat out wrong

So the prevalent idea in Australia public transport discourse that free/lower fares are “regressive” and bad for equality is simply wrong. And based on a simple logic error: Ignoring the distributional outcome of the tax that replaces the fares.

As far as I can tell, this notion seems to have been popularized in recent years in Australia mainly by RMIT Professor Jago Dodson, with output like the following:

Free public transport is a bad idea: RMIT expert

5 maps that show why free public transport benefits the affluent most

It’s sad and frustrating that this misguided thinking has taken off and been popularized when it entirely lacks substance. Jago Dodson should publicly withdraw his endorsement of these ideas. They are doing great harm to the cause of public transport and socioeconomic equality in Australia.

Public transport is about transport more than distribution anyway

While free/zero-fare public transport is a win for equality and poverty reduction, and we should do it on this basis, we ought to remember what the fundamental purpose of transportation systems is: Moving people around efficiently.

And what matters are overall outcomes. Not every single component of every single system needs to increase equality. Consider healthcare. Universal healthcare is good, in spite of mostly benefiting older people who are generally much wealthier than younger people, and thus being bad for equality. That’s no reason not to do it.

If you are concerned about inequality and poverty reduction (and we all should be), then the levers to pull are taxation, the welfare state, labour relations and economic policy, not public transport fares. Happily, free PT is good for equality. But even if it wasn’t, that wouldn’t necessarily mean we shouldn’t do it, because moving people around efficiently is the purpose and goal.

And so free public transport is good mainly because it is more efficient at moving people around! It gets people out of cars, it puts more money into services and infrastructure rather than fares, and it allows for quicker boarding/offboarding and quicker journeys with less hassle.

Free PT and EVs

The rise of EVs makes it particularly urgent in terms of getting people out of cars, or “mode shift” as they say in the biz, especially in a city like Melbourne where many people already own cars. This is because people make transport choices in terms of variable (ie per-trip) costs, not fixed costs.

EVs, unfortunately for people who want to see more PT use and less car use, have roughly zero variable costs, due to the often very low (or even negative) costs of charging. For public transport to compete on price, and not be relegated to a “luxury” option, it ought to be at least as cheap as a car for any given journey.

If EV journeys can cost roughly zero, that means PT not only ought to be free, it must be free if it is to remain competitive.