Registered unemployment — JSA claimant counts and rates — as far back as records allow
Years1971 – 2026
The headline rate measures JSA claimants — not unemployment.
The Isle of Man's published "registered unemployment" figure counts only people who are
actively claiming Job Seekers Allowance (JSA) from the Department of Health and Social
Security. As the IoM's own Labour Market Reports state: "there is no legal requirement
to register to receive Job Seekers Allowance." Separately, the IoM government publishes
an ILO-method estimate of total unemployment — in October 2025 this stood at
706 people, while the JSA claimant
count was approximately 261.
The ILO estimate is 2.7 times larger than the claimant count.
Roughly 445 people on the island were unemployed in October 2025 by any
reasonable definition — but were invisible to the headline statistic because they were not
claiming Job Seekers Allowance. The rate shown in this chart is the JSA claimant rate.
A COMIN press release citing "0.6% unemployment" is citing JSA claimants — a floor, not a ceiling.
Why so few people qualify for or claim JSA.
Job Seekers Allowance on the Isle of Man has two forms. Contribution-based JSA requires
sufficient National Insurance contributions in the previous two tax years — someone who
has been self-employed, worked part-time, or recently arrived on the island may not
qualify. Income-based JSA is means-tested: you and your partner must have less income
than the legally defined minimum, and savings, a working partner's earnings, or a
redundancy payment can all disqualify you. People who are between jobs, supported by
a partner, living off savings, or winding down a business will typically not qualify for
either form. This is not a failing of those individuals — it is a structural feature
of the benefit system that makes the claimant count a systematically narrow measure of
labour market distress, particularly in a high-income, high-savings, high-homeownership
island economy.
Series definitions
Isle of Man — three segments, one continuous series.
1975–1998 uses an ILO/World Bank national estimate as the best available proxy before a consistently published native series existed. From 1999, the series switches to the IoM's own published claimant rate, continued by Statistics IoM through to the present. These segments are the same underlying concept and the 1999 crossover is smooth; the Digest rates and the count-based rates from the most recent monthly table are consistent to within rounding. Two ILO estimate values were discarded as clear outliers against the native series: 2001 (ILO 1.6% vs native 0.49%) and 2006 (ILO 2.42% vs native 1.45%).
Source(s): ILOSTAT national estimate (World Bank/ILO, Dec 2019 vintage) for 1975–1998 · Isle of Man Digest of Economic and Social Statistics 2009, Table 3.6, for 1999–2008 · Isle of Man Labour Market Report, April 2026 ("Historic Isle of Man Registered Unemployment" table) for 2001–2026.
Guernsey — two measures with a substantial gap between them.
"Total registered unemployed" (the pre-2014 headline) includes people in part-time/casual work who remain eligible to claim Job Seekers Benefit as a supplement — this runs roughly 40–50% above the narrower "wholly unemployed" series. "Wholly unemployed" (the current headline, available from Q2 2011) counts only people with no work at all. Annual averages are means of four quarterly snapshots. Pre-1994 figures are approximate, reconstructed from narrative bulletin references rather than the spreadsheet itself.
Source(s): States of Guernsey supplementary data spreadsheet, Q4 2023 edition (quarterly snapshots Q1 1994–Q3 2024) · Guernsey Quarterly Population, Employment and Earnings Bulletin narrative references for pre-1994 estimates (Dec 1983 peak ≈770 people; Dec 1992 ≈760 people).
Jersey — "Actively Seeking Work" (ASW) registrants as % of workforce.
ASW includes people in partial employment seeking more work, making it closer in concept to Guernsey's "total registered" than to "wholly unemployed". Census ILO snapshots were materially higher than the registered rate at the same point in time — approximately 4.7% at the 2011 census and 3.5% at 2021, against ASW rates of around 2% and 1.5% — the gap is unregistered job-seekers invisible to the claimant count.
Source(s): States of Jersey press releases for 2001–2008 (approximate) · Statistics Jersey open data for 2009–2024 · Jersey Census 2011 and 2021 for the ILO comparison figures.
UK — claimant count rate, seasonally adjusted, all persons, annual average.
Before 2013 this counted Job Seekers Allowance claimants only — the closest like-for-like comparator to the island measures. From 2013 it also includes people claiming Universal Credit principally because they are out of work, as Universal Credit progressively replaced Job Seekers Allowance across Great Britain. By the time rollout was largely complete (early 2020s) this had widened the gap to the islands' JSA-only measures substantially: comparing the 2022–2025 UK rate (around 4%) to the IoM rate (under 1%) overstates the true difference in labour market distress, since a meaningful share of that gap is the broader Universal Credit eligibility test rather than a fourfold difference in actual joblessness. See the monthly tab for a fuller discussion, including why the Isle of Man's other benefits (Income Support, Employed Person's Allowance) don't provide a clean way to bridge this gap. Toggle off if distracting.
Source(s): Office for National Statistics, series BCJE, dataset UNEM (Claimant Count and vacancies time series), release 19 May 2026.
Which macro shocks actually show up in this data
The dotcom crash (2000–2002): no visible signal.
The IoM rate is essentially flat through this period (0.57% in 2000, dipping to 0.49% in 2001, 0.55% in 2002) and Guernsey shows the same pattern. This is consistent with what you'd expect: the dotcom crash was a NASDAQ-led collapse in technology and internet-equity valuations, and the Crown Dependencies' finance sectors at the time were weighted toward private banking, trust and fund administration, and insurance — not technology investment banking or venture capital. There's no economic mechanism connecting a US tech-stock crash to Manx or Channel Island employment, and the data bears that out. No event marker is shown for this on the chart, because there is nothing to mark.
Brexit (2016 referendum; 31 Jan 2020 exit; 31 Dec 2020 transition end): no visible signal either, and there's a clear reason why.
All three series are flat-to-falling through 2016–2019 and the only sharp move around the relevant dates is the March 2020 COVID spike, which swamps anything Brexit-related and arrives within weeks of the transition period actually beginning. The deeper reason no signal shows up is structural: the Crown Dependencies were never part of the EU and never benefited from UK membership, except for a narrow goods-trade protocol (Protocol 3), so EU exit didn't revoke a privilege they'd previously held. Their financial services market access into the EU runs through bilateral equivalence and third-country recognition rather than EU membership, and that didn't fundamentally change at the point of exit. The most-discussed practical friction was fisheries, which is economically marginal for all three islands. None of this means Brexit had zero effect — there may be slow-moving costs in areas like regulatory divergence, equivalence uncertainty, or firms' relocation decisions that don't show up as a unemployment-count shock — but if you were expecting a referendum-shaped dip or a 2020-transition-shaped dip distinguishable from COVID, it isn't there. No event marker is shown for the same reason as dotcom.
What the marked events represent, and what's deliberately left out.
The chart annotates only shocks with a clear, visible signature in the unemployment series itself: the early-1980s UK recession and the IoM's own 1985 peak, the early-1990s recession, the 2008 financial crisis, the IoM's own delayed 2013 claimant-count peak, and COVID in 2020. Shocks without a matching signal in this particular series — dotcom, Brexit, and for that matter the 2022–2023 inflation/rate-rise shock discussed in the inflation tab — are deliberately not marked here, even though they're real events, because adding a vertical line where the data shows nothing would imply a causal story that isn't supported.
The migration buffer — the most important structural factor.
All three islands operate flexible immigration systems that function as a labour market pressure valve in ways mainland Britain cannot replicate. A significant share of the employed workforce — particularly in construction, hospitality, and finance support roles — consists of people who came specifically for work and have no long-term attachment to the island. When work dries up, many leave. They exit the labour force entirely: no longer in the denominator of the unemployment rate, and certainly not registering for Job Seekers Allowance on their way out. The unemployment rate falls not because the economy has adjusted smoothly, but partly because the population has contracted to match available work. This is arguably the primary mechanism by which the islands absorb economic shocks, and it produces unemployment rates that systematically understate the true degree of labour market distress during downturns.
Source(s): analytical interpretation; no single published source, drawn from the general pattern of small-island migration economics and consistent with the population/workforce data discussed elsewhere in this observatory.
Recent IoM context — what the headline misses.
The JSA claimant count for the Isle of Man in recent years (2022–2025) is running at approximately 260–370 people per month — historic lows by any measure. But the ILO estimate published alongside the same report was 706 in October 2025, more than 2.7 times the JSA count. This gap — roughly 445 people — represents the unregistered unemployed: people who are out of work and looking but who either do not qualify for Job Seekers Allowance or have chosen not to claim it. This is the figure that matters if you are trying to understand actual labour market conditions rather than benefit caseload. Separately: the IoM aircraft registry has experienced significant contraction following sanctions-related deregistrations and exit of Russia/Belarus business; public sector departments have been managing budget pressure; and the Post Office has restructured. Whether these developments are visible in the claimant count depends largely on whether affected workers qualify for and claim JSA.
Source(s): Isle of Man Government unemployment statistics page (October 2025 ILO estimate and claimant count) · general reporting on aircraft registry, public sector, and Post Office developments, not independently re-verified for this note.
Show
These are JSA claimant counts — people actually claiming Job Seekers Allowance.
The counts shown are raw monthly end-of-period headcounts, not seasonally adjusted.
The COVID spike in April 2020 (1,347) reflects registrations under emergency schemes including
the Manx Earnings Replacement Allowance (MERA) alongside regular JSA — making it not directly
comparable to the surrounding series (shown in red).
Note also: the ILO estimate of total unemployment is consistently
2–3 times larger than these JSA counts.
The "Rate (%)" toggle divides the monthly count by an estimated workforce denominator
for that year — interpolated from the same annual rate/count pairs used in the annual
series tab — so treat the rate line as indicative rather than an officially published figure.
UK is now plotted in both modes: in "Count" mode it sits on its own right-hand axis (its
workforce is roughly a thousand times larger), and in "Rate (%)" mode it shares a single axis
with the IoM line, since both are percentages.
Source(s): "Historic Isle of Man Registered Unemployment" table, Isle of Man Labour Market Report, April 2026 (IoM monthly counts) · Office for National Statistics, series BCJD (count, thousands) and BCJE (rate, %), dataset UNEM, release 19 May 2026 (UK monthly figures).
Are IoM JSA and UK "JSA" actually the same thing? Mostly not, any more.
Both systems are nominally built on Job Seekers Allowance, and both have a contribution-based
form and a means-tested form, with broadly similar entitlement logic. But the comparison has
eroded badly since 2013. In Great Britain, JSA has been almost entirely replaced by Universal
Credit, and the UK figure plotted here explicitly includes Universal Credit claimants who are
principally out of work — a far broader and easier-to-qualify-for benefit than legacy JSA ever
was. The Isle of Man has no Universal Credit; its claimant count remains on the old, narrower
JSA-only basis throughout the whole series. From 2013 onwards, the two lines are not measuring
the same thing: the UK line increasingly reflects a different, more inclusive benefits regime,
while the IoM line stays on the stricter pre-2013-style test. Pre-2013, the comparison is much
closer to apples-to-apples (both JSA-only), so the earlier overlap in shape (2001–2012) is more
meaningful than the later one.
Could IoM Income Support be added to JSA to better match the UK's Universal Credit-inclusive figure? No — wrong population.
It's a reasonable instinct, since IoM does have Income Support, and UK's claimant count now
includes a broader UC caseload than legacy JSA alone. But the two benefits aren't filling the
same gap. IoM Income Support is specifically for people who are not required to be
available for work — its main groups are lone parents, people who are pregnant, those on
unpaid parental or paternity leave, and limited categories of students. It exists precisely
because these people are exempt from the work-seeking test that JSA requires. Adding Income
Support recipients to the JSA count would pull in carers and new parents, not job-seekers,
which would push the IoM figure in the wrong direction for this comparison — inflating it
with people who aren't unemployed by any working definition, rather than capturing people
who are out of work but ineligible for or not claiming JSA (the genuine undercount discussed
above, and the gap visible in the ILO estimate). The Isle of Man has no real equivalent to
Universal Credit's broader "in-work or out-of-work, single integrated benefit" design — the
closest analogue, Employed Person's Allowance, is itself restricted to people already working
a minimum number of hours, so it doesn't capture job-seekers either. There isn't a clean way
to construct an IoM aggregate that matches the UK's UC-inclusive scope, because the Isle of
Man's benefit system simply doesn't have a benefit that plays that role.
Source(s): Isle of Man Government, Income Support and Employed Person's Allowance guidance pages (gov.im) · UK Income Support and Universal Credit guidance, for the parallel UK definition (Income Support is for people not required to be available for work, and is separate from the claimant count in both jurisdictions).
A separate, smaller effect: IoM's wealth profile likely suppresses claims further.
Independent of benefit design, IoM's higher median income, savings, and home-ownership rates
plausibly mean fewer people fail the means test on the low-income side, and fewer bother
claiming a relatively small income-based payment in the first place if they have savings to
draw on. This would suppress the IoM claimant count somewhat below what an equivalent UK
population might show even under identical rules — though this is a smaller effect than the
JSA/Universal Credit definitional gap discussed above, and isn't separately quantified here.
What the monthly series shows.
The 2001–2008 period shows very low counts (150–750) with a clear trough in 2001 and a
sustained rise from mid-2003 through 2008 as the financial services sector experienced
structural pressure. The October–December 2008 uptick (reaching 721 in December) is the
first signal of the post-GFC deterioration, which peaked in January 2013 at 1,208 — the
highest point in this series. The recovery from 2014 to 2019 was steep, with counts falling
below 400 by late 2017 and reaching historic lows of around 290–310 in early 2018. The
COVID spike in spring 2020 was sharp but brief; by late 2021 counts were back below 300.
Since 2022 the series has been running at 240–380, broadly similar to the pre-2003 trough.
Seasonality.
A modest but consistent seasonal pattern is visible: counts tend to rise in November–February
and fall in May–September, consistent with the seasonal nature of hospitality, tourism, and
construction on the island. This is most clearly visible in the 2003–2008 and 2014–2019
sections of the series.
COVID anomaly (March–December 2020).
The spike from 389 in February to 850 in March and 1,347 in April 2020 reflects registrations
under the Manx Earnings Replacement Allowance (MERA) scheme as well as regular JSA.
MERA was a temporary COVID income support scheme; approximately 50% of unemployed survey
respondents in 2020 were on MERA versus 35% on JSA. The 2020 figures are therefore not
fully comparable to the rest of the series and will overstate the JSA-equivalent claimant
count for that period.
What the UK comparison shows, and what actually happens around 2016.
Both series share a similar broad shape through the GFC and the slow 2009–2013 grind — this
is the period where the comparison is most legitimate, since both were JSA-only measures.
The gap between them is tightest in 2015 (IoM 1.9%, UK 2.3%) and then widens steadily from
2017 onwards, but this isn't one shared event — it's two separate, genuinely real trends
crossing in opposite directions at almost the same moment. On the UK side, the claimant rate
troughs in 2016 (2.2%) and then turns upward through 2017–2019, partly reflecting the
maturing post-GFC cycle and partly the early effect of Universal Credit rollout broadening
who counts as a claimant. On the IoM side, 2015/16 was actually a weak year economically —
GDP fell 0.9% that financial year on declining e-gaming revenue — but this was followed by
a strong rebound: GDP grew 7.4% in real terms in 2016/17, with information technology,
e-gaming, and finance and business services cited as the main drivers. That rebound shows up
almost exactly in the claimant count, which falls from 1.55% in 2016 to 0.99% in 2017 and
keeps falling to 0.78% by 2019. So the widening gap is UK claimants rising for UK reasons
at the same time as IoM claimants falling for IoM reasons — not a single causal story, and
not primarily a Universal Credit artefact at this stage (that effect becomes more dominant
later, especially by 2022–2025, see below).
Source(s): Isle of Man Government, National Income Report commentary (gov.im, GDP growth and sector drivers for FY2015/16 and FY2016/17) · ONS series BCJE for the UK trough/rise pattern.
The later, larger gap (2022–2025) is a different story — mostly definitional.
From 2022 onwards the UK rate sits at roughly 4%, against an IoM rate below 1% — a gap of
this size is not plausible as a real difference in labour market distress between two
broadly comparable high-income economies, and should be read as reflecting benefit design
(full Universal Credit rollout by this point) and registration behaviour as much as economic
conditions. This is distinct from the 2016–2019 widening discussed above, which was driven
by real, opposite-direction economic trends on each side rather than benefit definitions.