In the song Big Yellow Taxi, Joni Mitchell beautifully encapsulated angst about the way we use land: “Don't it always seem to go, that you don't know what you've got till it's gone. They paved paradise and put up a parking lot.”This lyric resonates with many Ontarians. Residents of urban or suburban areas in the province—and that is most of the population—recall a different pattern of land use than what they see today. In some places, what used to be agricultural land is now used for other purposes. If we see these changes through the lens of Ms. Mitchell's lyrics, the apparent trend could be alarming. And when citizens become alarmed, government policy actions often follow.But Ontario is a big place, and what is happening to land use in urban areas may not be indicative of land use elsewhere in the province. In a recent study by the Fraser Institute, we examined the empirical evidence on the amount of agricultural land in Ontario from 1951 to 2013. What we found will confound those in the Big Yellow Taxi.There are two main measures of the amount of agricultural land area in Ontario: cropland area and farmland area. Cropland is land used to produce grains, oilseeds, hay and also vegetables and fruit. Farmland includes all of the cropland, but adds woodlots, wetlands, land used for farm buildings and some pasture. We think cropland is what most people have in mind when they think about agricultural land, and for this reason we think it is the more meaningful indicator. The area of cropland in Ontario actually increased slightly between 1951 and 2013, from about 3.5 million hectares to about 3.6 million hectares. Farmland area did indeed decrease during this time period, from about 8.4 million hectares to about 5.1 million hectares.But at the same time, farming efficiency has improved so much that less land is needed to produce more food. Consider that between 1981 and 2013, grain corn yields increased from about 6,000 kilograms per hectare per year to more than 10,000 kg per hectare per year. Winter wheat yields increased by 48 per cent and soybean yields increased by 40 per cent during the same period. As the amount of land required for agriculture shrinks due to higher productivity, some land probably should find its way to other uses, whether that's industrial, or recreational, or as parklands or whatever land-markets demand.And that's where the other interesting part of our findings come in: we looked at whether or not the land-use changes of the region were, or were not representative of “market failure.” Our findings, similar to previous researchers, suggest that the problem is more one of planning-failure than market failure. Measures to limit the amount of agricultural land used for non-agricultural purposes are based on a theory of “absolute advantage,” that the best agricultural land should be reserved for agriculture. But the best agricultural land is often the best land for use in other ways, and only markets, not planners, can determine this.Land-use planning can also lead to injustice. For example, the wide