A common view is that work is fundamentally a “good” thing. After all, if people don’t work, they’ll sit idle and not be able to provide for themselves. While the idea is simple enough, it has some powerful, albeit misleading, policy implications.

First, the accessibility of low-wage foreign workers through trade and globalization threatens the ability of Canadians and Americans to find work. President Trump has responded to these concerns by threatening to renegotiate the North American Free Trade Agreement (NAFTA).

Second, the “perfect” foreign worker that neither takes lunch breaks nor gets sick—automation—looms as a continual threat to employment. This week billionaire Mark Cuban warned Americans they should prepare themselves for massive job losses as machines take their jobs, while Bill Gates even advocated for a “robot tax.” Some, such as the CEO from Tesla, are beginning to suggest a “universal basic income” is the ultimate answer.

Finally, if work is a good, government has a mandate to help create it, through government programs, for example. In fact, one story suggests Obamacare has effectively created some three million jobs. Likewise, Prime Minister Trudeau recently earmarked millions of dollars for job-creation efforts.

Yet rather than address each of these implications, one might turn to their fundamental presupposition: do we really want work?

Consider, in the 17th to 19th centuries the average labourer worked some 60 or 70 hours a week. During the Middle Ages, labourers worked less formal work hours outside the household, but their days were chock full of backbreaking labour that was done inside the household, such as tending to animals, gathering wood, spinning yarn for clothes, and on it goes. Comparatively, we are all effectively unemployed—we work very, very little to sustain ourselves.

But are we really worse off than they?

Yale economist William Nordhaus found that, in terms of hours of labour needed to gain one hour of reading light, the average worker in 1800 needed to work about six hours (to obtain a tallow candle). Today, an hour of reading light requires a half a second of labour by the average worker (compact fluorescent light bulb). This time saved translates to time that can be used for other things, namely leisure and rest.

Likewise, according to U.S. statistics, buying a common basket of groceries (three pounds of tomatoes, a dozen eggs, one pound of bacon, and so forth) took 9.5 hours of work at the average U.S. wage rate and food prices of 1919. Today, it takes less than two hours. Again, this is time freed up to do other things.

To an economist, then, work is not a “good” but likely a “bad.” We don’t want employment, but rather the productive fruit that it brings. If work is a “bad,” the more productive we become, the less we want work. In fact, our desire is to become so productive that our “work” amounts to pressing a button in the morning, providing all the sustenance we require, which then frees us up to do whatever else we’d like to do—spend time with family, pursue hobbies, volunteer, or even perform research to satiate our curiosities.

The consequence is this: our main concern is not work, but productive capacity. If a machine can produce more for us, why would we want a human to do it? Do we want an iPhone, which even homeless people now enjoy, or a thousand people working to do in a year what an iPhone does in seconds? Our quality of life has risen precisely because machines have rendered many occupations obsolete.

As society grows in productivity, we all reap its rewards: workers aided by automation earn higher wages, factories produce complex goods at low cost, the poor obtain basic necessities with little toil, and yes, we all have more “unemployment” to enjoy our lives.

This piece was originally posted at the Fraser Institute.