A Ludicrously Oversimplified Explanation of the Canadian Economy Under COVID, the Increase in Debt and Money Supply, and What Comes Next — PART I

16 min readJun 22, 2020

TL;DR version: Due to fears of COVID and lack of childcare (rather than retail lockdowns), white-collar professionals (WCPs) have massively decreased their spending, causing parts of the economy to suffer a substantial reduction in income. Those sectors are being propped up through federal government transfers, financed by the Bank of Canada. What those WCPs do over the next few years will determine the state of the economy, inflation, the size of government interest payments, and social mobility.

What follows is what I hope we will be part of a series with 3 parts:

What’s going on with the Canadian economy, ignoring all international factors.
Now let’s examine the international economy.
What, exactly, is the Bank of Canada doing, and why?

I’ll largely be taking U.S. recent research by Raj Chetty (go read this piece at Vox if you haven’t already, then come back), massively oversimplifying it and Canadianizing the explanation.

There’s two important economic concepts that we need to understand before diving into what happened:

One person’s spending is another person’s income. For every thing sold, there must be a thing purchased. Or, put differently, one person’s expenditure is another person’s revenue. So if consumers are spending less, all else being equal, businesses have less revenue.
(Most) economic activities are only captured in the data when there is a financial transaction. If I pay a company to mow my lawn, my expenditures have gone up, and the landscaping company’s revenues have gone up, and there is a flow of payments to workers, governments (through taxes, etc.). If instead, I mow my own lawn, there is no change in any economic statistics. Either way, the grass has been cut, but in the second case the value of the cut lawn, the effort exerted, etc. does not appear anywhere. That’s because our standard economic statistics ignore “home production”, which is problematic for all kinds of reasons, particularly because it undervalues the contribution of women to the economy, as they engage in far more home production (a.k.a. unpaid labour) than men.

That in mind, let’s get on to the analysis.

Point 1: Fear of COVID, and need to look after children, keep people, particularly white-collar professionals at home. Retail closings only play a supporting role.

Data from the United States shows that timings of “closings” of the economy and “reopenings” explain surprisingly little about the path of economy activity. The U.S. provides a fascinating natural experiment, as U.S. states closed and re-opened at different times, and yet there was very little difference in economic activity between those states:

The explanation here is straight-forward: what happens in the economy is due to the decisions made by millions of individual households and businesses. This can absolutely be influenced by governments, so I’m not suggesting that lockdowns aren’t playing a role (they absolutely are). But before the closings even hit, workers (particularly white-collar professionals) were finding ways to work from home, either because they were afraid of the virus, or because they were pulling their kids from schools and daycares, and needed to be at the house.

Not everybody has the capability to work from home. It is largely white collar professionals that have jobs that can be performed at home, rather than at an office, and have the disposable income to shop online, rather than at brick and mortar stores. It is their actions that are driving the economy.

Point 2: White-collar professionals are spending less and performing more home production. This is driving the economic downturn.

I have to admit, I audibly gasped when I saw this figure in Chetty’s research:

Source. Vox. Yes, I really did gasp. Yes, I’m a nerd.

That “Top Income Quartile” of households are largely those white-collar professionals (WCPs) mentioned in Point 1. While lower-income households did reduce their spending somewhat when the crisis hit, their spending is roughly back where it was pre-COVID. The top 25% of households, however, are still spending 1.4 billion USD less, per day, than they did in 2019. We don’t have data for Canada, but given the sizes of the economies, we would expect it to be somewhere in the 100–200 million CAD per day range.

In the U.S. data, we do see some rebound in the spending of WCPs. You would think (hope) that this would be happening even faster in Canada, given our COVID infection rates are significantly lower than in the United States. Over the next few months, we’ll learn about the extent of WCP spending rebounds, as Statcan releases data on retail sales, etc. I’m less optimistic than some that we’ll see a particularly large Canadian bounceback, because there’s still large parts of the economy that are inoperable (daycares), and what drives WCP behaviour is not the virus, but fears of the virus, and Canadian WCPs are likely influenced by what they see happening in the United States.

We do have some Canadian data from TD Bank, on spending by consumers and businesses (unfortunately not broken down by income quartile). Like the United States, the overall consumer spending gap is closing, but spending is still 4% lower than the year before. Canadian business spending is still well below 2019 levels, which is not promising.

Anyhow, it’s important to note that WCPs are not spending less because they’re earning less income; although there has been some job loss among this cohort (as well as people cutting their hours in order to take care of their kids), it pales in comparison to the job loss experienced by lower-income households, who have hardly adjusted their spending at all. (Paul Jacobson has some interesting data on reductions of hours worked by province and educational attainment, showing that university graduates have relatively modest drops as compared to other Canadians, with men being far less affected than women). Rather they are earning about the same amount of money, but spending less. The U.S. Bureau of Economic Analysis is showing an unprecedented spike in household savings rates:

Source: CNBC. If anyone has good Canadian data, please send it my way.

A Canadian sign that net savings are up, is that credit card expenditures aredown dramatically and debit purchases are up. This is an indication that consumers have extra money in their bank accounts, so they don’t need to finance their purchases with credit, but can pay with (digital) cash.

But spending is not down uniformly across the economy. Some sectors of the economy are doing quite well, including food delivery and video streaming, as shown by this data from April:

But, overall, the sales accruing to the winning sectors is far smaller than the sales lost by the losing sectors, as spending is down overall. (For the time being, I won’t differentiate between the businesses in the sector and the people who work for those businesses.)

The shift to working at home (and the kids staying home) is a big part of this. Daycare expenditures are way down, but so are any expenditures involving being at work, from transportation costs to purchases of coffee near workplaces. Starbucks is closing a couple hundred stores in Canada; I’d bet that almost all of these are in downtown locations. With office buildings half open for the foreseeable future, sales at these stores will be way down, so it’s probably not necessary to have dozens of them in a one-mile radius.

Expect many of these to be empty by the end of the year. That’s going to create some interesting challenges for commercial real estate.

Both of our economic concepts from the beginning of the piece are vital here:

One person’s spending is another person’s income. For every thing sold, there must be a thing purchased. The reduction in spending of white-collar professionals has drastically reduced incomes in the losing sectors, with a less-than-offsetting increase in the income of winning sectors, causing an overall drop in economic activity.
(Most) economic activities are only captured in the data when there is a financial transaction. This shift in economic activity is not a free-lunch for white-collar professionals (WCPs). Coffee and meals are still being made, kids are still being looked after, etc. However, instead of paying service workers to do these tasks, WCPs are doing these tasks themselves through home production. This home production does not show up in the economic statistics anywhere, due to a lack of financial transactions, but it absolutely is work. It just is work that does not raise someone else’s income, or does not generate tax revenues for governments. The workload of WCPs has increased tremendously, and although their market incomes have not risen, their savings have, as their work has displaced expenditures. This has equity ramifications, as this work is disproportionately carried out by women. (And too many men are gleefully unaware of this, even in their personal lives, as the New York Times headline Nearly Half of Men Say They Do Most of the Home Schooling. 3 Percent of Women Agree shows.

The shift from market production to home production has also significantly reduced tax revenues for governments, leading to higher deficits (and debt). Municipalities are hit particularly hard, as one big losing sector is public transit, so they get dinged twice, through both reduced tax revenue (e.g. property tax revenue from closing retail locations) and through a drop in income.

In the absence of any interventions by governments or the Bank of Canada, we can summarize the economic impacts by a chart. I can’t stress this enough: this is a ridiculously oversimplified analysis. We’re ignoring any international factors, we’re lumping both workers and businesses into winning sectors and losing sectors, and we’re ignoring large parts of the economy, like the financial sector.

Outcomes in the absence of government assistance

It’s important to note that incomes and expenses have fallen by exactly the same amount here (7 arrows) — that’s not a coincidence, as one person’s spending (expense) is another person’s income. As well, net savings in the economy is unchanged. WCPs and winning sectors are saving more, losing sectors and governments are saving less.

But this analysis is for the hypothetical world where government transfers didn’t change. What happens if we add them in?

Point 3: Government transfers to losing sectors is preventing an even bigger collapse. This is replacing the lost spending from WCPs, and will continue until spending by WCPs recover.

The Parliamentary Budget Office (PBO) has put out financial estimates for the federal government for the 2020–21 fiscal year. It shows a reduction in tax revenue, along with an increase in Employment Insurance spending (which would occur naturally without government intervention). The biggest increase, however, is in “other transfer payments”, which are an alphabet soup of programs designed to prevent financial distress for workers and businesses:

This spending is sometimes characterized as “stimulus spending”, and it certainly has a stimulus effect, but that is not the primary purpose of these programs. Rather, it is to avoid a wave of personal and business bankruptcies and avoid mass financial hardship among those in losing sectors.

This will continue in some form, it has to continue in some form, so long as WCPs continue to replace expenditures with home production. The only way to get the economy to recover is to create the conditions for WCPs to return to something resembling their former levels of expenditures, which at the very least involves creating the conditions where they are comfortable (and able) to put their kids back into schools and daycares. This is likely going to take a very long time, and involve a whole lot of stops and starts, as John McNally and I describe in We aren’t restarting the economy. We’re climbing a mountain.

I’m not by nature an optimist, so I expect this period of depressed economic activity and high government deficits to persist for some time. I hope I’m wrong.

Adding government transfers to our chart, noting that those transfers are not perfectly targeted toward losing sectors, we have the following:

Three things to note about this chart:

Given the size and speed of government transfers, it was impossible to perfectly target them. As such, not all have made their way to losing sectors (and given the reduction of jobs and hours among some WCPs, it’s wholly appropriate that those individuals would receive some supports.)
Although the intent of these policies was to keep losing sectors solvent, this did have a stimulus effect — market incomes are up (or less low than they otherwise would be) in both winning and losing sectors.
This stimulus impact has reduced the overall economic hit. Our economy has now declined by 5 arrows, rather than 7.

But we still have this issue of increased government debts. For every seller of debt, there must be a buyer. So who is buying this stuff?

Point 4: The Bank of Canada, rather than the private sector, is financing this debt.

Later in this series, I’ll get into the mechanics of what the Bank of Canada is doing (and why). But their activities are relevant to this discussion, so here’s a high level discussion.

An obvious follow-up question to Point 3 is: “How is the government financing all of these transfers and this (projected) 256 billion dollar deficit?” Even Andrew Scheer is asking it.

The PBO has a fantastic FAQ on the federal debt. It includes the following projections for government debt:

Note the increase in “unmatured debt”, projected to hit 1 trillion CAD by the end of the fiscal year. The current (as of the end of April) level of market debt is just over 900 billion; with nearly 60% of this maturing in the next 3 years:

The PBO notes that federal issuances of treasury bills (short term debt instruments that mature within a year or less) increased “from $127.1 billion on February 29, 2020 to $258.0 billion on April 30, 2020”.

So who is buying all of these bonds and T-bills? Turns out, it’s primarily the Bank of Canada. Here is how their holdings of each have evolved since 1995:

An increase in government bond holdings by the Bank is not unprecedented, as they substantially increased their holdings in the late 1990s, and again in the early 2010s. But the level and speed is something Canada has never experienced. Here’s the chart, starting at the beginning of the year.

That is a 95 billion dollar increase in Treasury Bill holdings, along with a 75 billion dollar increase in Bond holdings. Combined, that’s an amount roughly equal to the increase in T-Bills and Bonds issued by the Government of Canada.

Given that the Bank of Canada is a crown corporation owned by the federal government, this debt is money that the federal government owes to itself, with one part of the federal government paying interest to another part of the federal government. So, for the time being, this debt is essentially costless for the government to finance. However, as we will discuss, this could change.

I’ve (intentionally) used “savings” rather sloppily in this piece until now. But moving forward, I’ll replace “savings” with “money” and “fed bonds” (which I’ll use to refer to both federal government bonds and T-bills). The government has increased the supply of bonds, and the Bank of Canada has purchased those bonds by increasing the supply of money. This leaves us with the following:

Details about this chart:

I’ve marked the Bank of Canada’s “selling” of money and the federal government’s selling of federal bonds in red, as these were not pre-existing assets, rather they have been created during this crisis.
The overall supply of both money and federal bonds has increased, as shown on the far-right column.
For every bond sold, or money “sold”, there has been a buyer. In the case of fed bonds, the Bank of Canada has purchased an amount that is roughly equivalent to the amount created by the federal government.
In the case of money, the newly created money by the Bank of Canada, has largely flowed to WCPs and winning sectors, who have increased their savings, while losing sectors have seen a decrease in their monetary holdings.
This increase in money has not caused an increase in inflation (in fact, inflation rates have fallen during this crisis, though COVID-related changes in purchasing patterns are distorting inflation data). The easiest way to think about inflation is “too much money chasing too few goods”, but WCPs aren’t using that money to chase goods (remember, their expenditures are down substantially). This has caused a decrease in the velocity of money (the speed at which money propagates through the economy). Below is data from the U.S.; monetary velocity has been dropping for the past 20 years. It fell substantially during the financial crisis, and looks to be again during this one.

All bad things must come to an end. What happens when this crisis is over, and we still have all this created money and government debt?

Point 5: The Bank of Canada may be able to monetize all of this debt forever, causing no rise in federal interest payments. That’s not necessarily a great outcome.

This crisis is going to have all kinds of permanent effects. I particularly worry about the scarring effect the recession will have on new graduates, who are starting their careers in an economy with few jobs and little possibility for advancement. But for this piece, I will only focus on the long-run impacts of this increase in debt and the money supply.

The conventional wisdom about what happens next looks something like follows:

When this crisis is over, there will be significant pent up demand by WCPs, who will go out and spend down their savings.
This increase in spending will boost the economy (yay!), but will also create significant inflationary pressures. Too much money chasing too few goods, and all that.
To tamp down this inflation, the Bank of Canada will be forced to sell their bond holdings and reduce the supply of money. In order to incent the general public (or whoever) to buy these bonds, interest rates will have to go up.
This will be effective, and the rate of consumption growth will slow, as people are buying fewer cars and more bonds.
The federal government will get dinged twice here. First, they’ll get dinged because rather than the Bank of Canada holding these bonds, non-governmental actors will be. So instead of paying interest to themselves (that is, paying interest to the Bank of Canada, which then flows back to the government), they’ll be paying interest to someone else in the economy, causing their net interest payments to rise.
The federal government also gets dinged because interest rates have risen, so as all of those T-bills and bonds rollover (remember, 60% of them mature within 3 years), the government will have to finance this debt at higher interest rates, which also causes their net interest payments to rise.
This increase in interest payments forces the federal government to cut spending (or slow their rate of spending growth) or raise taxes.

That’s an outcome we should be worried about, for sure. And it certainly might happen. But there are other potential outcomes we need to concern ourselves with. Here’s another:

When this crisis is over, WCPs go back to spending about what they did before, and there’s not a significant level of “pent up demand”.
Economic growth remains tepid and inflation largely does not materialize.
Interest rates stay low, this growth in the money supply is permanent, and the Bank of Canada continues to hang on to this debt. The debt is effectively monetized.
Federal government interest payments stay around pre-crisis levels, and there is no pressing need to cut spending or raise taxes.
WCPs use their increase in savings in a number of ways. They use it to pay down consumer and mortgage debt (which puts further downward pressure on interest rates). They sink it into the stock market; this is already happening, with a boom in day-trading during this crisis. They use it to purchase even more residential real estate.
Although consumer inflation remains low, this increase in the money supply drives asset price inflation, particularly in the stock and real-estate markets. We’re already seeing this in the stock market, as the influx of day traders is pushing stock prices up. The S&P 500 is down only about 8% from its February 2020 peak, despite the world being in the midst of a massive economic crisis.
WCPs may also gift the money to their kids, so their kids can buy real-estate. This further pushes up real estate prices in our cities, and effectively locks-out any young person that does not receive this wealth from their parents. This causes a reduction in social mobility, as kids from lower-income households are locked out from the most economically dynamic areas. This gifting of cash also allows the kids of WCPs to take unpaid internships, which gives them a leg up on their career. None of this would be new, but rather an acceleration of trends that we’ve seen over the past couple of decades.

I’m not in the predictions game — I’m not sure what will happen. What I can say is that there are many possible futures, and we need to be ready for all of them. In general, I don’t think Canadian policy wonks have thought through the implications of a rise in savings among WCPs, and what happens to the Canadian economy if this money isn’t used for consumer spending any time soon.

I hope you found this useful. I plan on updating it based on your questions, so please send them my way!

Update: A few hours I posted this, the Bank of Canada published Targeting inflation during the pandemic. It’s a must read. I’m pleasantly surprised how similar this piece is to the Bank of Canada’s thinking.