Building · Metric · Entrepreneurship & the digital economy

Business entry and exit in international context

Canada creates businesses barely faster than it loses them. For every business that exited in 2021, 1.13 entered — down from 1.48 in 1984. In 2023 the entry rate fell to 12.3% while the exit rate rose to 11.9%, a percentage point above the record low set the year before. Canada still runs ahead of the United States on this measure, which matters more than it sounds: the United States is the only G7 country whose business-exit data can be put on the same axis as ours.

Business formation is the entry point to every other metric in this pillar. A country that is not starting firms has nothing to take public, nothing to patent, and nothing to scale. Canada measures this unusually well — Statistics Canada runs three separate programmes on business death alone, one of them monthly — and the picture they return is not the one the monthly headlines suggest.

Before the numbers: is a high exit rate actually bad? No — and this page does not argue that it is. Business exit is how capital, premises, and workers get released to more productive uses; an economy with no exits is an economy where nothing new can get a foothold. That is precisely why the headline here is the ratio of entry to exit rather than the exit rate on its own. Canada's exit rate has fallen over forty years. The problem is that its entry rate has fallen further and faster. Low churn in both directions is sclerosis, not stability.
Exit rate · 2023
11.9%
Up 1.1 percentage points in a single year, from 10.8% in 2022 — the lowest level on record. An exit is permanent by construction: no employees this year and none the next.
Businesses entering per business exiting · 2021
1.13
1.48 in 1984. The ratio has fallen below 1.00 — more exits than entries — three times in the published record: 1991, 2013, and 2020.
Bankruptcies as a share of businesses · 2004–2020 average
0.22%
Against 9.0% of businesses exiting annually over the same period. Bankruptcy is roughly one-fortieth of business failure, and it is the only part with a 1987 time series.

Entry still outruns exit, by a margin that has narrowed for forty years

Statistics Canada's own long-run work puts Canada and the United States on one chart, which settles the comparability question for the only peer where it can be settled. The measure is the ratio of the entry rate to the exit rate: above 1.00, more businesses are being created than lost; below 1.00, the business population is shrinking.

Canada entered the 1980s creating roughly three businesses for every two that closed. It now creates about nine for every eight. The decline is not a Canadian peculiarity. The American series falls by exactly as much over the same span — 0.35 on the ratio, from 1.37 to 1.02, against Canada's 1.48 to 1.13 — so Canada has kept its lead without arresting its own slide. The pandemic year is visible and exceptional in both series, and in opposite directions: Canada's ratio collapsed to 0.82 in 2020, the lowest value in the published record, while the American ratio held at 1.02.

Businesses entering per business exiting, business sector, 1984–2021 (entry rate ÷ exit rate; 1.00 = one entry per exit)
Source: Statistics Canada, “The long-run evolution of business entry and exit rates in Canada”, Economic and Social Reports, catalogue 36-28-0001, vol. 5 no. 9 (2025009), article 00001, Chart 2 and its published data table. Canada from the Longitudinal Employment Analysis Program (1984–2000) and the National Accounts Longitudinal Microdata File (2001–2021); United States from the US Census Bureau Business Dynamics Statistics. Series ends 2021, the latest year published on this basis.

The levels behind the ratio move the same way. Earlier research quoted in the same article, covering 1983 to 2012, found Canada's entry rate fell from 24.5% to 13.1% while the exit rate fell from 16.5% to 11.6% — entry down by close to half, exit down by under a third. Statistics Canada's latest annual figures show the gap closing again from the other side: entry at 12.3% in 2023, down from 13.9%, and exit at 11.9%, up from 10.8%.

Bankruptcy is the tip of the iceberg, and it is the part with the longest history

Almost every public discussion of business failure in Canada runs on insolvency filings, because the Office of the Superintendent of Bankruptcy has published them monthly since 1987 and they arrive within about six weeks. They are also the smallest possible slice of the phenomenon. Statistics Canada linked its business microdata to the insolvency register and found that over 2004 to 2020, an average of 0.22% of incorporated employer businesses filed for bankruptcy each year and 0.04% filed a proposal to creditors, against 9.0% that exited. Nearly all bankruptcies end in exit; almost no exits involve a bankruptcy.

Share of incorporated employer businesses per year, annual average 2004–2020 (percent)
Source: Statistics Canada, “A profile of corporate exits and insolvencies”, Economic and Social Reports, catalogue 36-28-0001, vol. 3 no. 10 (2023010), article 00005, 25 October 2023. Population: incorporated businesses with employees, excluding educational services, health care and social assistance, and public administration. Bankruptcies and proposals are filings under the Bankruptcy and Insolvency Act; filings under the Companies’ Creditors Arrangement Act and receiverships are excluded.

The practical consequence is that a rise or fall in insolvency filings is close to useless as a read on business failure, and that the only measures worth benchmarking are the ones that count businesses ceasing to have employees, whether or not a court was ever involved.

Four of the G7 can be compared with each other. Canada is not one of them.

There is no harmonised business-closure indicator for the G7. Two incompatible definitions split the group. Canada and the United States count a business as closed when it stops having payroll employment, with no waiting period. The United Kingdom, France, Germany, and Italy count an enterprise as dead when it is de-registered from the business register, and only if it has not been reactivated within two years. The two bases differ in level, in timing, and in what they are actually measuring; putting them on one axis would produce a ranking that means nothing.

The register-basis group can be ranked against each other, and the spread within it is wide: France at 10.93% and the United Kingdom at 10.8% against Italy at 5.71%, all for 2023 and all provisional, with the European Union average at 8.51%. Canada, the United States, and Japan publish nothing on this basis at all.

Enterprise death rate, business economy, 2023, provisional (percent of active enterprises). No bar means no series is published on this basis
Sources: Eurostat, dataset bd_size, indicator ENT_DTHR_PC “Death rate - enterprise deaths divided by active enterprises - percentage”, all enterprises, NACE B–S excluding O and S94, dataset updated 13 March 2026 (France, Germany, Italy, and the EU27 average of 8.51%); Office for National Statistics, “Business demography, UK: 2024”, 20 November 2025, Table 1 (United Kingdom; the 2024 provisional rate is 9.8%). Canada, the United States, and Japan publish no register-basis enterprise death rate — see the table below for what they do publish.

Japan is the binding constraint on any G7 ranking of business exit. Its only official figure is an exit rate of 3.9% for fiscal 2023, derived from employment-insurance records rather than from a business register or a payroll-transition rule — a third basis again, and not comparable with either group. The timelier Japanese numbers, from Teikoku Databank and Tokyo Shoko Research, are commercial products rather than official statistics.

CountryWhat is publishedBasisFrequencyLatest
🇨🇦 CanadaMonthly openings and closures; annual exitsPayroll employmentMonthly and annualClosures 4.8% (Dec 2025); exits 11.9% (2023)
🇺🇸 United StatesEstablishment closings, and establishment deathsPayroll employmentQuarterlyClosings 5.2%; deaths 3.0% (Q1 2025)
🇬🇧 United KingdomEnterprise deaths; a quarterly closure count in developmentBusiness registerAnnual and quarterly9.8% (2024, provisional)
🇯🇵 JapanExit rate from employment-insurance recordsInsured establishmentsAnnual, fiscal year3.9% (FY2023)
🇫🇷 FranceEnterprise deathsBusiness registerAnnual10.93% (2023, provisional)
🇩🇪 GermanyEnterprise deathsBusiness registerAnnual8.87% (2023, provisional)
🇮🇹 ItalyEnterprise deathsBusiness registerAnnual5.71% (2023, estimated)

Canada’s two entries are not interchangeable: a monthly closure rate and an annual exit rate measure different events over different windows. Sources and definitions for every row are in the companion workbook.

The monthly series is the timeliest in the G7 and the most widely misread

Canada is the only G7 country publishing a monthly business closure figure. Table 33-10-0270-01 has run since January 2015, arrives about three months in arrears, and is built from Canada Revenue Agency payroll-deduction remittances. Statistics Canada still labels it experimental.

It is also routinely read as a count of businesses lost, which it is not. A closure in this series means a business had employees last month and none this month; it may be seasonal, temporary, or a restructuring. Statistics Canada is explicit that “a business can close and reopen within a year, [while] a business can only exit or enter once.” The releases prove the point by decomposing the opening rate: in December 2025, reopenings ran at 3.2% against new entries at 1.6% — two-thirds of that month’s openings were businesses coming back rather than businesses starting.

Reference monthOpening rateClosure rateof which new entriesof which reopenings
July 20255.0%5.0%not publishednot published
August 20254.7%4.8%not published3.3%
October 20254.5%4.9%1.7%2.9%
December 20254.8%4.8%1.6%3.2%

Source: Statistics Canada table 33-10-0270-01, seasonally adjusted, as published in each reference month’s release of The Daily. Only months whose rate levels appear in a release are shown; September and November 2025 and January to April 2026 are omitted because those releases do not publish the levels. Entries and reopenings are rounded independently of the opening rate, so the components need not sum to it.

Where the series does earn its keep is direction. Through 2025 the closure rate ran consistently at or above the opening rate, and the active business population fell in most months — small changes, but in one direction. Statistics Canada reports that the deviation is above the 2015-to-2019 norm on both sides: in December 2025 the opening rate stood 0.1 percentage points above its historical average and the closure rate 0.2 points above.

Findings

Finding 1

The margin between creation and destruction has thinned to almost nothing

Canada created 1.48 businesses per exit in 1984 and 1.13 in 2021. In 2023 the entry rate was 12.3% against an exit rate of 11.9% — a gap of 0.4 percentage points, where in 2022 it had been 3.1. A business population can shrink without any single dramatic year.

Finding 2

Both rates fell, but entry fell faster — that is the finding, not the exits

Over 1983 to 2012 the entry rate fell from 24.5% to 13.1%, close to a halving, while the exit rate fell from 16.5% to 11.6%. Fewer firms are dying than a generation ago. Far fewer are being born.

Finding 3

Canada is ahead of its only comparable peer

Canada's entry-to-exit ratio was 1.13 in 2021 against 1.02 for the United States, and Canada's entry rate is higher than the American one in every year of the published record, at the precision the source reports. On the one comparison that is methodologically clean, Canada does not look bad — it looks like a country declining from a better starting point.

Finding 4

The number everyone quotes is the wrong number

Bankruptcies averaged 0.22% of businesses a year against 9.0% exiting, so insolvency filings capture roughly one exit in forty. They are quoted because they are monthly and go back to 1987. Business closure and business bankruptcy are not the same measurement, and only one of them describes the economy.

Sources & methodology

Primary sources
Methodology notes
  1. Three measures, kept apart. A closure is a business with payroll employment last month and none this month; it may be temporary. An exit requires no employees in the reference year and none in the following year, which makes it permanent by construction and puts the series roughly two years in arrears. A bankruptcy is a legal filing. The three differ by orders of magnitude and are never mixed on this page.
  2. Why the headline is a ratio. An exit rate read alone cannot distinguish a dynamic economy from a failing one. The ratio of entry to exit can: it falls when creation slows relative to destruction, whichever of the two is moving. It is also the one long-run series Statistics Canada publishes for both Canada and the United States on a consistent basis.
  3. Comparability across the G7. Canada and the United States use a payroll-employment definition with no reactivation window; the United Kingdom, France, Germany, and Italy use business-register de-registration confirmed over two years; Japan uses employment-insurance establishment records. These are three different measurements and are never ranked against each other here. The register-basis countries are charted together; Canada, the United States, and Japan appear as absent series rather than as invented values.
  4. A precision limit worth naming. The long-run article publishes its entry and exit rate series to two decimals as proportions (0.24, 0.16) under a heading reading “percent”, which resolves a rate only to about half a percentage point. Those levels are recorded in the companion workbook for transparency and are deliberately not charted. The ratio series is published to two decimals as a ratio and is precise; it is what the chart uses.
  5. Figures computed rather than published. Three: the 0.4 and 3.1 percentage-point gaps between the entry and exit rates in 2023 and 2022; the statement that bankruptcies are about one-fortieth of exits (0.22% against 9.0%, a ratio of 40.9); and the observation that reopenings were two-thirds of openings in December 2025 (3.2% of 4.8%). Each is arithmetic on two published figures, and each is shown as such in the workbook.
  6. What is missing and why. A continuous annual entry and exit series for 2001 to 2023 is not reproduced: table 33-10-0164-01 is served through a JavaScript interface with no fetchable data file, and the releases of The Daily publish only the most recent year or two as levels. The monthly table is shown only for the months whose levels appear in a release. Missing data is left missing.
  7. Companion workbook. The editable series live in Building_Business-Entry-and-Exit_Data.xlsx, which carries the full 1984–2021 ratio series, the rate levels as published, the annual and monthly Canadian figures, the exit-versus-bankruptcy comparison, the register-basis death rates, the United States quarterly detail, and a G7 comparability sheet recording what each country publishes.
Why this metric sits under Building, not Elevating

Business failure could be read as a household-hardship measure, which would place it with the citizen-outcome metrics. It is not treated that way here. Entry and exit describe the economy's capacity to form and reallocate productive capital — the same question as the startup pipeline, venture capital per capita, and intellectual property generation, all of which sit under Building. The pillar asks whether Canada can still build; the rate at which Canadians start firms, and the rate at which those firms survive, is the most direct available answer.

It also belongs with those metrics for a practical reason: they share a failure mode. Each measures a stage of the same pipeline, and each can look acceptable on its own while the pipeline as a whole is thinning. Read together, the four make a claim that none of them makes alone.

Page last reviewed August 2026 · Data current to 2023 — the latest published by Statistics Canada