Undercover Economist

The window tax — an open and shut case

‘People respond in profound ways to tax incentives. They adjust their behaviour to avoid tax’

The adage ‘free as air’ has become obsolete by Act of Parliament,” thundered Charles Dickens in 1850. “Neither air nor light have been free since the imposition of the window tax. We are obliged to pay for what nature lavishly supplies to all, at so much per window per year; and the poor who cannot afford the expense are stinted in two of the most urgent necessities of life.”

Dickens prevailed: the window tax, which had been levied in England since 1696, was abolished within a year. But the curious story of the tax, explored recently by Wallace Oates and Robert Schwab in the Journal of Economic Perspectives, holds lessons for us today.

The details of the tax varied across the centuries but with the broad theme that the more windows your house had, the more tax you had to pay. At first glance, the tax seems clever, even brilliant. Rich people had larger houses, and so paid more tax. Windows are easy to count from outside the premises, so the tax was easy to assess. Poor people didn’t own large houses, so they weren’t affected by the tax. And the number of windows in a house doesn’t change, so the tax was impossible to avoid.

Wrong, wrong, wrong.

The tax was probably progressive but not nearly as progressive as it might seem. Many poor people did live in large houses — as servants or in tenement blocks. They suffered from the tax, as we shall see. Adam Smith himself, in The Wealth of Nations (1776), nailed the other problem with the idea that the tax was paid only by the rich: “A house of £10 rent in the country may have more windows than a house of £500 rent in London.”

A more fundamental error is the idea that architecture doesn’t respond to tax incentives. When William Pitt tripled the tax in 1797, thousands of windows were bricked or boarded up almost overnight. Later, the president of the society of carpenters in London told Parliament that almost every homeowner on Compton Street had approached him to reduce the number of windows. A new apartment building in Edinburgh was designed with an entire second floor filled with windowless bedrooms.

When Dickens complained that the poor were being denied light and air, he wasn’t speaking figuratively. Poor people did not have to pay the tax out of their own pockets but their landlords did, and the poor dwelt in stuffy darkness as a result.

After 1747, the window tax followed a strange structure. Houses with fewer than 10 windows paid no window tax; those with 10-14 windows paid six pence per window per year. As a result, the cost of having a 10th window was that you also had to pay tax on the other nine. Tax wonks call such discontinuities “notches”, and there were further notches at 15 and 20 windows.

If these notches seem absurd to you, modern governments don’t seem to have a problem with them. Stamp duty, a tax on property transactions in England and Wales, contained notches until last year, and the UK income tax system recently acquired a new notch: a transferable tax allowance for married couples, worth more than £200, evaporates abruptly if one of the couple strays into a higher tax bracket, even by a single pound. All of this distorts our behaviour.

Nonsensical as the notches are, they help economists to see the effect of the tax. In the recent study, Oates and Schwab combed through tax records from the mid-1700s. They found that nearly 50 per cent of the total number of houses had the tax-efficient totals of nine, 14 or 19 windows — an intelligent response to a foolish tax. People who bricked up a couple of windows to bring the total down to nine were inconvenienced by the tax, yet the Treasury earned no revenue from them. Most taxes will produce some of this sort of waste but the window tax was particularly egregious.

If it seems strange that tax policy could shape the architecture of a country, consider New Orleans’s distinctive camelback houses, one storey high at the front (the part of the home that’s taxable) but with two storeys at the back — a tax-efficient architectural style.

And ponder the research of economists Joshua Gans and Andrew Leigh, who noted that twice as many births were recorded in Australia on July 1 2004 than on June 30 2004. Why? The July babies were eligible for a “baby bonus” of A$3,000 and the June babies were not. Gans and Leigh even found that many Australians delayed their deaths — or perhaps the moment their deaths were recorded — long enough to escape inheritance tax when it was abolished on July 1 1979. If our births and deaths respond to tax incentives, it shouldn’t be surprising that a few windows might be bricked up.

There is a useful lesson to be learnt from the window tax: it is that people will respond in quite profound ways to tax incentives. That is why economists often call, more in hope than expectation, for a tax on carbon emissions. People would adjust their behaviour to avoid the tax, which is exactly what we need.

But perhaps a more realistic lesson is this: it’s perfectly possible for a bad tax to last for 155 years.

Written for and first published at ft.com.