Junk Economics - SUPER QSIX: The Great Canadian Bean Boom

By Bryan Moir | Junk Economics

This morning I harvested beans.

Quite a lot of beans, actually.

Three containers now sit on the kitchen table overflowing with purple and green beans, accompanied by several cucumbers and one rather lonely tomato apparently included for statistical diversification.

I harvested beans last week too.

But looking only at this morning’s photograph, I could announce something considerably more exciting:

Production is booming.

Output has accelerated.

The garden is outperforming expectations.

The fundamentals remain strong.

We have momentum.

Put an upward-pointing arrow beside the photograph and I have myself a perfectly serviceable government economic release.

There is only one small problem.

You don’t know how many beans I harvested last week.

You don’t know how many plants produced them. You don’t know whether the harvesting intervals were identical. You don’t know whether this is simply the normal August peak. You don’t know cumulative production. You don’t know yield per plant. You don’t know how this harvest compares with last year’s.

And you certainly don’t know whether the plants are entering their most productive period or approaching the end of it.

All you know is that there are a lot of beans sitting on my table.

The beans are real.

The narrative is the invention.

And that brings us rather neatly to Canada’s economy.

Statistics Canada reported that employment increased by 75,000 in July, or 0.4%. The unemployment rate declined to 6.4%, while the employment rate increased to 60.9%.

That followed gains of 88,000 jobs in May and 18,000 in June. Canada therefore added about 181,000 jobs over those three months.

Those are good numbers.

There is absolutely no reason to pretend otherwise.

In fact, doing so would commit precisely the sin this column exists to expose: deciding what the statistics should say before examining what they actually say.

But 181,000 jobs sitting in a statistical bowl do not prove an economic renaissance any more than three bowls of beans prove an agricultural one.

They require context.

So let’s QSix the harvest.

Employment has rebounded strongly since April.

But the choice of starting date matters.

Statistics Canada reported that May’s 88,000-job increase was the first significant employment gain since November 2025. More importantly, it followed a net employment decline of 112,000 during the first four months of 2026.

April itself recorded another 18,000 decline in employment, while unemployment rose to 6.9%.

Then:

May: +88,000

June: +18,000

July: +75,000

Photograph only those three bowls and the garden looks spectacular.

Extend the camera slightly to the left and we discover that some of what we are observing is recovery from previous weakness.

That doesn’t make the recovery imaginary.

It changes what we can legitimately infer from it.

There is an enormous analytical difference between:

things are getting better

and

things have never been better.

There is an equally important difference between:

employment has rebounded

and

Canada has entered a new high-growth economic trajectory.

The first proposition is supported by the recent data.

The second requires considerably more evidence.

This may be the most dangerous question in economics.

It ruins speeches.

Canada’s employment rate reached 60.9% in July.

Sounds good.

Now compare it with the recent past.

The employment rate was 60.8% in January 2026.

It was 61.1% in January 2025.

And 61.8% in January 2024.

Suddenly the photograph changes.

The July employment rate represents an improvement from the spring weakness, certainly.

But it is hardly evidence of an unprecedented employment boom.

Indeed, the overall unemployment rate of 6.4% remains above its 2017–2019 average of 6.0%. Core-age unemployment is 5.5%, compared with a pre-pandemic average of 5.1%, while youth unemployment is 12.6%, compared with 10.8%.

Again:

Compared with what?

Suppose I announce that my bean harvest increased 40%.

Impressive.

Then I tell you I planted 70% more bean plants.

Less impressive.

Economic quantities without denominators make wonderful politics.

They make lousy economics.

Now empty the bowls onto the table.

One of the most persistent errors in public economic discussion is treating every additional job as economically interchangeable with every other additional job.

They are not.

A job is obviously valuable to the person who gets it.

But if we are trying to determine whether the productive capacity of an entire economy is improving, composition matters.

Full-time and part-time work are not identical measures of labour utilization. Market-sector and government-sector employment tell us different things about private economic activity. Employment gains in highly productive, capital-intensive industries are economically different from identical numerical gains in lower-productivity activities.

That does not make one worker more worthy than another.

It means GDP accounting is not a moral ranking system.

The July number tells us employment increased.

QSIX asks the next question:

Where?

Because an economy can add tens of thousands of jobs while simultaneously losing productive capacity elsewhere.

The aggregate doesn’t necessarily reveal the structure underneath it.

My garden could produce 200 additional beans while my cucumber crop collapses.

Announce only TOTAL VEGETABLE PRODUCTION UP and the cucumbers disappear.

Nothing false has been said.

Something important has simply been omitted.

Welcome to bumper-sticker economics.

The recent monthly GDP numbers provide another attractive bowl.

Real GDP by industry expanded 0.3% in May, with both goods-producing and services-producing industries contributing to the increase.

That is unquestionably positive.

Now widen the frame.

Real GDP was unchanged in the first quarter of 2026, following a 0.2% decline in the fourth quarter of 2025. Final domestic demand edged down 0.1%.

Business capital investment fell 0.7% in the first quarter.

That was its fifth consecutive quarterly decline.

Residential investment fell 2.0%.

Now our booming economy has become rather more complicated.

May’s GDP increase is still real.

It simply no longer carries the evidentiary burden originally assigned to it.

This distinction matters enormously.

A monthly acceleration after a period of stagnation may indicate that an economy is turning a corner.

It does not prove that the road beyond the corner is straight, paved and heading uphill.

Sometimes there is another corner.

Economists invented time series because photographs were insufficient.

Now we arrive at the statistic that economic cheerleading generally keeps somewhere behind the curtains.

Productivity.

Canadian business labour productivity fell 0.5% in the first quarter of 2026, following a 0.3% decline in the previous quarter.

That matters far more to Canada’s long-term prosperity than whether one particular month produced 75,000 additional jobs.

Because employment measures how many people are working.

Productivity asks what we are producing with the labour we use.

And Statistics Canada’s own comparison with the United States is brutal.

From the first quarter of 2015 through the third quarter of 2025, Canadian labour productivity grew at an annualized rate of only 0.80%.

The United States managed 1.96%.

Canadian real GDP per capita grew 0.82% annually over that period.

The United States: 2.07%.

Statistics Canada concludes that Canada’s productivity performance has failed to keep pace with the United States and identifies the relative decline in productivity growth as the principal reason Canada has been falling behind across broader measures of economic performance.

Those aren’t numbers produced by some disgruntled economist blogging from his garden.

They come from Statistics Canada.

The bean equivalent is straightforward.

Suppose I harvest 50% more beans this year.

Excellent.

Now suppose I needed twice as many plants, twice as much fertilizer and twice as many hours of labour to produce them.

Have I become more productive?

No.

I have become bigger.

There is a difference.

And Canada has spent far too much time congratulating itself for the former while neglecting the latter.

Eventually every economic discussion has to arrive here.

Not:

Did GDP increase?

Not:

Did employment increase?

Not:

Did some financial-market index increase?

But:

Statistics Canada’s Canada–U.S. comparison gives us a disturbing longer-term answer.

Between 1997 and 2015, Canadian and American real GDP per capita grew at almost precisely the same annualized rate: 1.53% in Canada and 1.52% in the United States.

After 2015 the trajectories separated.

From 2015 through the third quarter of 2025, Canadian real GDP per capita grew only 0.82% annually, while the United States grew 2.07%. By Q3 2025, Canada’s real GDP per capita relative to the United States was about 14% below its early-1997 relative level.

That is the structural problem against which July’s employment report must be judged.

Meanwhile, prices have not politely returned to where they were before the inflation surge.

Canada’s CPI was still 2.8% higher year-over-year in June, while grocery prices were up 3.9%. Grocery inflation had exceeded headline CPI inflation for 17 consecutive months.

And here we encounter another favourite statistical parlour trick.

When inflation falls, prices do not generally fall.

They rise more slowly.

If something costs $2, rises to $4 and then its inflation rate falls to zero, it still costs $4.

Nobody sends the missing two dollars back.

Declining inflation is good.

Pretending that declining inflation reverses the previous increase in the price level is nonsense.

Once again:

the statistic can be true while the story attached to it is wrong.

Yes.

That shouldn’t be difficult to say.

Recent employment numbers have been strong. The unemployment rate has declined. The employment rate has improved from its spring weakness. Monthly GDP strengthened in April and May.

Those are favourable developments.

If they persist, they will become evidence of something more important.

But they have not yet erased the structural evidence sitting beside them:

Canada entered the recent employment rebound after losing jobs during the first four months of the year.

The employment rate remains below its January 2025 level.

Business capital investment has declined for five consecutive quarters.

Business labour productivity fell again in Q1.

And Canada’s post-2015 productivity and GDP-per-capita performance has badly lagged the United States.

None of those facts disproves July’s 75,000 jobs.

They answer a different question.

And that is precisely the point.

Which brings me back to my kitchen table.

There really are a lot of beans there.

They aren’t imaginary beans.

Nobody manipulated the bean statistics.

Statistics Canada did not secretly adjust my purple beans into green ones.

There is no clandestine network of Marxist horticulturalists operating out of Agriculture Canada.

I had a good harvest.

What I cannot legitimately tell you from today’s photograph is that bean production is accelerating.

I cannot tell you productivity is rising.

I cannot tell you this year’s total harvest will exceed last year’s.

I cannot tell you next week’s harvest will be larger.

And I certainly cannot announce a permanent transformation of the Moir agricultural economy.

To establish those propositions, we would need more information.

We would need a baseline.

A denominator.

A time series.

Inputs.

Outputs.

Seasonality.

Historical comparisons.

And evidence that whatever improvement we are observing can continue.

In other words:

we would have to do economics.

That is the difference between reporting economic statistics and using economic statistics to tell a political story.

The first asks:

What does the number say?

The second too often asks:

What can I make the number say?

Those are not the same exercise.

The July employment report is good news.

Three months of improving labour-market data are better than three months of deterioration.

But neither a politician nor a newspaper headline gets to turn recovery into renaissance merely by changing the adjective.

Three bowls of beans do not prove an agricultural boom.

Three months of stronger employment do not prove an economic miracle.

For now they establish something considerably more modest:

Things may be getting better.

Keep measuring.

And whenever somebody tries to tell you considerably more than the data can support, there is one question worth asking before all the others: