Do construction companies survive better or worse than the UK average?

The question behind the insolvency headlines

Construction has the highest count of annual company insolvencies of any UK sector, and that number has risen roughly 50% over the past decade.^[2] It would be reasonable to conclude from that alone that a new construction business has a worse chance of survival than a business started in almost any other industry. The Office for National Statistics publishes a separate dataset that lets you check that conclusion directly, and the answer is not what the insolvency headlines imply.

ONS Business Demography tracks cohorts of newly born enterprises, construction included, and follows what percentage of each cohort is still trading one, two, three, four and five years after birth.^[1] We have compiled the full construction-versus-all-industries series in our UK Construction Survival Index. This article works through what it shows, and reconciles it with the insolvency data covered in our companion article on construction insolvency risk.

The headline number: construction outlasts the UK average

Of the construction enterprises born in 2019, 43.3% were still trading five years later, at the end of 2024. Across all UK industries, the equivalent figure was 38.4%. That is a 4.9 percentage point advantage for construction, on the most recent birth cohort for which a full five-year figure exists.^[1]

Put another way: a construction business born in 2019 was roughly one-eighth more likely to still be trading in 2024 than a newly born business drawn from the UK economy as a whole. That is not a marginal rounding difference. It is a consistent, measurable gap that shows up across every cohort in the dataset once enough years have passed to compare.

The full cohort data, year by year

ONS Business Demography Table 4.2 tracks five birth cohorts with at least one year of survival data available in the current release: 2019 through 2023. Not every cohort has reached every survival year yet, since a cohort born in 2023 cannot have a five-year figure until 2028. The table below sets out every year of data currently available for each cohort, construction against all industries.^[1]

Birth year	Measure	Year 1	Year 2	Year 3	Year 4	Year 5
2019	Construction	93.7%	75.6%	60.0%	50.1%	43.3%
	All industries	94.6%	74.7%	55.9%	45.0%	38.4%
2020	Construction	93.7%	74.2%	58.2%	49.5%	n/a
	All industries	92.9%	71.3%	53.0%	44.0%	n/a
2021	Construction	94.3%	73.3%	57.7%	n/a	n/a
	All industries	93.4%	70.4%	53.5%	n/a	n/a
2022	Construction	93.5%	69.0%	n/a	n/a	n/a
	All industries	92.3%	68.9%	n/a	n/a	n/a
2023	Construction	93.5%	n/a	n/a	n/a	n/a
	All industries	93.4%	n/a	n/a	n/a	n/a

Reading down each column shows the pattern clearly. Year 1 survival is close between construction and all industries in every cohort, generally within about one percentage point, and the direction is not always the same: construction ran slightly behind the all-industries rate in the 2019 cohort (93.7% against 94.6%) but slightly ahead in 2020, 2021, 2022 and 2023. From year 2 onward, construction is ahead in every cohort with data, and the gap tends to grow as the cohort ages: in the 2019 cohort it moves from a 0.9 point deficit at year 1, to a 0.9 point lead at year 2, a 4.1 point lead at year 3, a 5.1 point lead at year 4, settling at 4.9 points by year 5.

The one point in the data that does not fit the widening pattern cleanly is the 2022 cohort at year 2, where construction (69.0%) and all industries (68.9%) are effectively level. Whether that cohort widens again at year 3, in line with every earlier cohort, will only be confirmed once ONS publishes its next annual update. We treat that as an open question rather than force it into the general pattern, and the live figures are kept current at our UK Construction Survival Index.

Why this does not contradict the insolvency numbers

Construction company insolvencies rose from 2,793 in 2016 to 4,188 in 2025, a 49.9% increase over the decade, and creditors voluntary liquidation now accounts for roughly three in four of those cases.^[2] That is a real, well-documented trend, and it sits alongside the survival data above without contradicting it, because the two data sets measure entirely different things.

• Survival data tracks a birth cohort. It takes every construction enterprise born in a specific calendar year and asks what fraction is still trading one, two, three, four and five years later. It says nothing about companies that were already ten or twenty years old when the clock started.
• Insolvency data tracks annual events across the whole stock. The 4,188 construction insolvencies registered in 2025 were drawn from the entire population of trading construction companies, regardless of age, including long-established main contractors carrying large multi-year contracts, extensive payment chains, and material balance-sheet exposure to cost inflation and late payment.
• The two measures respond to different pressures. A newly born business typically starts small, with limited fixed contract exposure and lower overhead, and its main survival risk is the ordinary attrition every start-up faces (undercapitalisation, lack of a client base, founder decisions to cease trading). An established contractor's insolvency risk is driven more by fixed-price contract exposure, retention timing, and the payment behaviour of clients and main contractors above them in the chain, exactly the mechanics covered in our insolvency risk guide.

A sector can therefore show both patterns at once: newly formed businesses that survive their first five years at a better-than-average rate, and a rising annual count of insolvencies concentrated in the broader, ageing, more contractually exposed stock of trading companies. The construction data does exactly that. Neither number is wrong. They are answers to different questions.

What "survival" actually measures

ONS defines a business birth as a new enterprise registering for VAT or PAYE in a given year. Critically, this definition is not limited to companies incorporated at Companies House. It includes sole traders and partnerships that register for VAT or PAYE, which makes the survival dataset broader, and arguably more representative of the real population of CIS subcontractors, than either the Companies House incorporation data in our UK Construction Index or the Insolvency Service data in our UK Construction Insolvency Index, both of which only capture limited companies.^[1]

"Survival" in this data means the enterprise is still active and trading, not that it is profitable, growing, or free of financial stress. A construction business can survive five years while barely breaking even, changing legal form partway through (for example, a sole trader incorporating, which ONS treats carefully to avoid double-counting), or scaling down substantially from its first year of trading. The data answers a narrower question than "is the business doing well": it answers "did the business still exist and trade at each anniversary."

ONS also control-rounds all counts to the base 5 and does not publish a survival year until enough calendar time has genuinely elapsed since the cohort's birth year, which is why the 2020 cohort has no year-5 figure yet, the 2021 cohort has no year-4, and so on. We have not estimated or interpolated the missing years; they are shown as not yet available in the table above, consistent with how ONS itself presents the series.^[1]

What this means for a trade business owner

Two working conclusions follow from the data, and both matter for how a trade owner reads their own risk exposure.

First, starting a construction business is not, on this evidence, an unusually bad bet relative to starting a business in another UK sector. The common narrative, that construction is uniquely failure-prone from day one, is not supported by the ONS survival series. If anything, the five-year data for the most complete cohort available (2019) shows the opposite: a construction start-up had close to a one-in-eight better chance of reaching its fifth anniversary than the average new UK business.

Second, the real risk that shows up so clearly in the insolvency numbers is not primarily a start-up risk. It is concentrated in the ongoing operating exposure that comes with holding fixed-price contracts, extended payment terms, and retentions, the mechanics that determine whether an established trading business can absorb a bad debt or a cost shock without failing. That is a risk a business faces throughout its life, not disproportionately in its first year, and it is largely independent of whether the business happens to be new or long-established. A subcontractor's actual defences against it (short payment terms, credit checks on contractors, a diversified client base, and a company structure that keeps CIS cash flowing through the monthly Employer Payment Summary rather than an annual Self Assessment wait) are the same regardless of how many years the business has been trading. Our guide to CIS sole trader versus limited company and our overview of gross payment status and cash-flow resilience both work through those defences in detail.

Read together with the incorporation and insolvency picture in our state of UK construction business 2026 article, the survival data completes a fuller picture: construction is a sector with high formation volumes, better-than-average long-run survival for the businesses that are formed, and a real, rising rate of insolvency concentrated in the ongoing operating risk that every established contractor and subcontractor carries. Planning for the long term in a trade business means taking both parts of that picture seriously, not just the part that makes the headlines.

The full cohort series, updated as ONS publishes new Business Demography releases, is available at our UK Construction Survival Index.

1. Office for National Statistics, "Business Demography, UK: Table 4.2, Survival of newly born enterprises, broad industry group," release covering data to 2024, published 20 November 2025. Figures compiled in our UK Construction Survival Index. Data sourced under the Open Government Licence v3.0.
2. The Insolvency Service, "Company Insolvency Statistics -- Record-Level Data for England, Wales and Scotland," data through May 2026, published 19 June 2026. Figures compiled in our UK Construction Insolvency Index. Data sourced under the Open Government Licence v3.0.