The statistics are shocking.  Multiply the size of the average B.C. farm (365 acres) by the average cost per acre (\$5,321) and you get an average farm value of \$1.9 million.  Divide that by the average net profit for B.C. farmers (\$26,000) and you discover that, it takes 75 years for an average farm to recoup the cost of its land — and that’s if it spends every penny of profit on the land and the mortgage is interest-free.  As my pitch video for The Lands that Feed Us1 loudly states, “the average farmer will be dead by the time the average farm is paid for.”

It’s very headline worthy.  It’s also not very good information.  To be sure, all the stats are accurate (for the year 2016) — they come from Statistics Canada.  But the picture they paint is … misleading.  The problem is averages.  An average is a way of distilling a single number from a bunch of different data points.  It’s a way to make the raw numbers understandable.  When the data points are all represent variations of the same situation, this is helpful:  An average gives us a sense of what the situation is like in general.

But when the data points represent different things, taking an average is misleading because it deliberately obscures the differences.  When the data lumps together two (or more) situations that are different, the average is representative of neither.  This is what’s going on in the headline-worthy conclusion I drew at the start.  Farm and farmers in B.C. are not a monolithic group, and their situation changes depending on where they are in the province and what kind of farming they do.

Let’s take the numbers one by one.  We’ll start with farm size.  The average farm size in B.C. is 365 acres.  That number comes from the 2016 Census of Agriculture.  It’s calculated by dividing the total area of farmland (6.4 million acres) by the number of farms (17,528).  The question is, what counts as a farm?  According to the census, 1.4 million of those 6.4 million acres were in active use as “cropland”.  The remaining 5 million acres were pasture, woodland, wetland, or fallow.  In other words, the majority of the land counted as farmland was not cultivated.  If it’s used for growing food at all, it’s pasture for raising animals (likely cattle), but much seems to be either unused or unusable.

What about the 1.4 million acres of cropland?  60% is devoted to growing hay, and an additional 33.7% were being used for “field crops”, of which the majority comprised of wheat, canola, and oats.  4.4% was fruits, berries, or nuts, and only 1.1% was vegetables.

The point of all this statistical trivia is that no farm in B.C. is 80% pasture, 12% hay, 6% wheat and so on.  Rather, we have a bunch of different types of farms, all of which operate on different scales with different land requirements.  But the statistics lump them all together.  Because pasture and, to a lesser extent, hay require the most land, they dominate the picture.  If we break down the “average” 365 acre farm by land use, we discover that 285 acres of that average comes from land that is uncultivated.  Even hay, the largest cultivated crop would only represent 48 acres of this mythical “average farm”.  And vegetables — the type of crop that the public would most likely associate with farming — represents a miniscule 0.88 acres of the average.

Because so many different farm types are lumped together, the average isn’t very representative.  If you want to know the average size of a range in B.C., 365 acres is probably in the right ballpark.  But if you want to know how big our vegetable farms are, 365 acres is a ludicrous estimate.  At a guess (since we don’t actually have the number) it’s probably much closer to 10.

Next number:  Cost per acre.  The \$5,321 per acre average in B.C. is quite expensive compared to the national average (\$2,696), but if you try to buy land at that price in the Fraser Valley, you won’t get far.  Anecdotally, the market price is currently \$100,000 to \$300,000 per acre.  Of course, if you are buying a range in 100 Mile House, you’ll get a lot closer to the \$5,000 average.  And if you are buying in the Okanagan, you’ll probably pay five figures per acre.  Once again, the average is skewed by the vast pasture lands that make up most of the land area but which represent only one specific situation.  If you’re a young farmer looking for land for a vegetable farm, basing your business plan on the Stats Canada average land value is not a good idea.

The average net profit per farmer — \$26,000 — was a bit of a shot in the dark.  I estimated this number by dividing the total net cash income for B.C. farmers by the number of census farms.  Surely this number, derived from income tax data, would not suffer the same “false average” issues as the other numbers.  Perhaps not.  But what counts as a farmer?  Stats Canada tracks income for both “census farms” and “farm operators” — which one is more representative of “B.C. farmers”?  Which definition you choose makes a \$10,000 difference — almost 40%.

Another concern is whether “net cash income” accurately represents farmer income at all.  Net income is a simple number:  It’s money received minus money spent, but it doesn’t include costs that aren’t directly related to growing the food — and land, along with buildings and machinery, is one of those costs.  For the purposes of my example this was useful, since I wanted to know how much money was available to pay for land, and I didn’t want to double count the cost of land.  But, it also meant that I over-estimated the amount of profits that the “average” farmer took home.

“Realized net income” from the same Stats Canada data table might arguably have been a better estimate of the amount of money that farmers took home at the end of the day, since this would have accounted for those extra costs.  But this number has its own issue:  2015 is by far the highest year (at least between 2013 to 2017), which means the 2015 “average” isn’t really representative of normal farmer profits.  “Realized net income” (totalled for all farmers) ranged from -\$139 million in 2013 to \$52 million in 2015, and three of the five sample years were negative; farmers lost money in those years. I don’t have enough background in statistics to do a proper confidence interval, but if we look at the actual variation — \$52 million ± \$191 million — it’s very clear that the number changes far too much from year to year to be reliable.  The variation is almost four times bigger than the number we’d be citing!  Ironically, this is a situation where taking an average would have helped:  Averaging the numbers over multiple years would have smoothed out the year-to-year differences and yielded a number that was more representative of farmers’ financial situation.

Ok, let’s come up for air.  I hope I’ve done a convincing job of destroying the credibility of the stats I used in my pitch video.  Now I want to talk about why I used the numbers the way that I did.

In one sense, the story I tell — that the average farmer will be dead by the time the average farm is paid for — is a classic example of lying with statistics.  I made a claim, found some stats to support it, and bingo, instant credibility!  Factually, it’s all true, the numbers are real, and the conclusion I draw does follow from the data I based it on.  It’s headline worthy, and I sum it up in 15 words.  That’s a lot less than the 1,200 or so words I just wrote to explain why the “average farmer” doesn’t mean much.

In another sense, despite all my bellyaching, there is something fundamentally out of whack with the cost of farmland.  What I’m saying isn’t all that crazy.  If you strip away the “average” gobbledygook and look at some of the specific situations, it’s true that farmers can’t afford to buy land.  A small farmer in the Lower Mainland does have to pay six figures per acre, they will need at least a few acres to make a go of it, and, if they do very well, their income will be somewhere in the low to mid five figures.  Those are deliberately vague numbers to illustrate how the argument looks without statistics, but you can tell that the numbers don’t add up just by looking at the orders of magnitude involved.

So, why use statistics?  Why deliberately tell a headline story when I know the statistics aren’t good enough?

I have a guilty admission to make:  I’m selling something.  Remember how I said the numbers came from a pitch video?  The pitch video is aimed at a \$50,000 grant from Telus that will help fund this project.  I want that pitch to be successful.  And, if I’ve learned anything about pitching for funding, it’s that the pitch has to be a) quick, b) attention-getting, c) easy to grasp, and d) a compelling story.  Those are the criteria on which the pitch will be judged — not factual accuracy or depth of understanding.  The headline — “The average farmer will be dead by the time the average farm is paid for” — that’s the hook, designed for easy consumption by people who have no background knowledge of the issue.

The hope is that, if I can get attention — and funding — with the headline, I can convince people to slow down enough to pay attention to The Hands that Feed Us.  And my goal for The Hands that Feed Us is to tell the complicated story:  To show what it’s like to make a living as farmer, and to dig into all the specifics that are erased by averages and statistics.  To show that, even if it’s true that farmers don’t earn enough to pay for their land, somehow, we still have farmers.

1. The Lands that Feed Us is a spin-off from The Hands that Feed Us that will look at farmland access issues in B.C.