The conventional story

The standard narrative of UK economic history goes something like this: steady post-war growth, punctuated by crises — the IMF bailout of 1976, the recession of the early 1980s, the financial crisis of 2008. Plot annual GDP growth rates and you get a noisy line oscillating around a long-run mean, with a sharp dip in 2008–09 that the economy eventually recovered from.

The conventional conclusion: volatile but mean-reverting. A story of shocks and recoveries.

But we have been asking the wrong question. The question is not “did growth temporarily fall?” It is “has the economy’s underlying trajectory permanently shifted?” These are completely different questions — and they require completely different analytical methods.

Part 1 — The wrong question: annual growth rates

Here is what the three most common chart types tell you when applied to UK annual GDP growth rates (%) from 1949 to 2025.

Chart 1: X-mR Shewhart Control Chart — Annual Growth Rate %

The X-mR chart is the most commonly used SPC chart in quality improvement and process monitoring. Applied to UK annual GDP growth rates:

X-mR Control Chart of UK Annual GDP Growth Rate
X-mR Control Chart — UK Annual GDP Growth Rate (%). One stage detected. The 2008 financial crisis and 2020 Covid contraction are the only signals — everything else is classified as common cause variation.
What the X-mR tells you: The UK economy is a stable process with a mean growth rate of 2.43%. Two exceptional events — 2008 and 2020 — stand out. Otherwise: business as usual for 77 years.

Chart 2: Run Chart — Annual Growth Rate %

Run Chart of UK Annual GDP Growth Rate
Run Chart — UK Annual GDP Growth Rate (%). The median line sits at 2.43%. Only the 2008–2010 recession generates a meaningful run signal below the median.
What the Run Chart tells you: Growth has been broadly stable around the median, with a notable dip around 2008–2010. Perhaps something changed recently — it’s hard to say with any precision.

Chart 3: Bootstrap CUSUM — Annual Growth Rate % — Result: 1 Stage

Now apply the most sophisticated method — Bootstrap CUSUM step-change analysis — specifically designed to detect structural change:

Bootstrap CUSUM of UK Annual GDP Growth Rate — 1 stage detected
Bootstrap CUSUM — UK Annual GDP Growth Rate (%), 90% confidence, 3,000 loops. Result: 1 distinct stage. No structural change detected in 77 years of annual growth rate data.
Even Bootstrap CUSUM finds nothing in the annual growth rate data. At 90% confidence with 3,000 bootstrap resamples, the method correctly reports: 1 distinct stage. No permanent structural change detected. Three charts. Three methods. Same conclusion: nothing to see here.

And yet the UK economy has been visibly, structurally slowing for six decades. How can three legitimate analytical methods all miss it? Because they are all answering the wrong question.

Part 2 — The right question: cumulative level

Annual GDP growth rates measure year-to-year fluctuation. They are mean-reverting by design — recessions cause temporary dips; recoveries bring them back. The right question is not about the acceleration. It is about the trajectory.

Cumulative GDP level data — built by compounding annual growth rates into an index starting at 100 in 1949 — captures that trajectory directly. When Bootstrap CUSUM is applied to cumulative level data, it answers: “has the economy’s long-run growth path permanently shifted?”

Chart 4: X-mR Chart — Cumulative Index

X-mR Control Chart of UK Cumulative GDP Index
X-mR Control Chart — UK GDP Cumulative Index (1949=100). The control limits, calculated from the overall mean of ~206, are meaningless for a series that starts at 100 and ends at 285. Almost every observation from 1985 onwards breaches the upper limit.
What the X-mR tells you about the cumulative data: Every observation from 1985 onwards is a “signal.” The chart has broken down entirely — it cannot handle trending data because it assumes a stable process mean.

Chart 5: Run Chart — Cumulative Index

Run Chart of UK Cumulative GDP Index — median at 206, 1987 turning point visible
Run Chart — UK GDP Cumulative Index (1949=100). Every observation before approximately 1987 is below the median (blue); every observation after is above (red). The 1987 turning point is visible — but the run chart draws one flat median through what is clearly a climbing staircase.
What the Run Chart tells you about the cumulative data: Something changed around 1987. But not how many changes occurred, when each happened, or with what confidence.

Chart 6: Bootstrap CUSUM — Cumulative Index — 90% Confidence — 8 Stages

Bootstrap CUSUM of UK Cumulative GDP Index at 90% confidence — 8 distinct stages
Bootstrap CUSUM — UK GDP Cumulative Index (1949=100), 90% confidence, 3,000 loops. Result: 8 distinct stages. The staircase of structural growth regime changes is clearly visible across 77 years.

Notice the CUSUM turning point — the peak of the green CUSUM line — occurs around 1987. This is the moment at which accumulated deviation from the long-run mean was greatest. After 1987, the cumulative index began consistently exceeding its historical average. Bootstrap CUSUM identifies this as the boundary between Stage 5 and Stage 6 — the shift from the Lawson boom era into the New Labour stability period.

Same data. Completely different picture. Applied to annual growth rates: 1 stage, nothing detected.
Applied to cumulative level: 8 distinct stages at 90% confidence. A continuous staircase of structural shifts.

What the data actually shows

Bootstrap CUSUM Stage Analysis — UK Cumulative GDP Index (1949 = 100), 90% Confidence
StagePeriodIndex MeanChange %Growth Regime
11949–1958115.35BaselinePost-war recovery
21958–1966143.19+24.1%Macmillan boom
31966–1976172.55+20.5%Wilson turbulence, sterling crisis
41976–1986194.16+12.5%IMF crisis, end of post-war consensus
51986–1996220.04+13.3%Lawson boom, financial deregulation
61996–2003241.90+9.9%New Labour stability
72003–2014261.77+8.2%Pre-crisis peak — absorbs 2008 entirely
82014–2025277.40+6.0%Post-austerity new normal — weakest in 77 years
Finding 1 — The 2008 Financial Crisis was not a structural break Stage 7 runs from 2003 to 2014, absorbing the 2008 Financial Crisis entirely. Bootstrap CUSUM finds no evidence that it created a new structural growth trajectory.
Finding 2 — The real structural breaks were elsewhere Stage boundaries fall in 1958, 1966, 1976, 1986, 1996, 2003, and 2014 — at moments of institutional and policy change, not at crisis points.
Finding 3 — The long deceleration
+24.1% → +20.5% → +12.5% → +13.3% → +9.9% → +8.2% → +6.0%

Each successive stage delivers less growth than the last — a 60-year structural deceleration, continuous across governments of every political persuasion.

Note: observe how the steps in the Bootstrap CUSUM chart get smaller over time — each stage boundary is closer to the last than the one before it. This is the visual signature of structural deceleration: the economy is still growing, but each growth regime delivers less than its predecessor.

Does the conclusion hold at higher confidence levels?

Bootstrap CUSUM at 95% confidence — 5 stages
95% confidence — 5 stages. Three marginal stages drop out but the staircase and deceleration remain clearly visible.
Bootstrap CUSUM at 99.7% confidence — 4 stages
99.7% confidence — 4 stages. Only the four largest structural shifts survive at the most conservative threshold. The deceleration from +24% to +6% is still unambiguous.
The core finding is robust across all confidence levels. Whether you choose 90%, 95%, or 99.7%, the same story emerges: a structural deceleration in UK growth that predates 2008, Brexit, and Covid by decades. This robustness is itself a form of evidence.

What this means for economic analysis

If the deceleration predates 2008, 2016, and 2020, then fixing those “events” will not reverse it. The structural breaks identified — 1966, 1976, 1986, 2014 — suggest trajectory shifted at moments of institutional and policy change. Understanding why those breaks occurred is a more useful question than relitigating the financial crisis or Brexit.

⚠ Honest caveats

  • This uses total GDP, not per capita GDP — population growth affects the picture
  • GDP measurement methodology has changed over 77 years
  • The analysis identifies that shifts occurred and when — attribution requires separate economic analysis
  • This is a statistical description, not an economic forecast

Try it yourself

📋 Step by step — UK GDP cumulative index

  1. Download UK annual GDP growth data from the World Bank or ONS — or use the prepared CSV download above
  2. In Excel, create a cumulative column: set 1949 = 100, then =prev_row * (1 + growth_rate/100)
  3. Add a Year column: 1949, 1950, 1951…
  4. Save as CSV and upload to stepchangeanalysis.com
  5. X-Axis = Year, Y-Axis = Cumulative, Date Format = YYYY, Turn Length = 10, Loops = 3000, Conf % = 90
  6. Click Recalculate Chart

Run this analysis on any time series data

Free, browser-based, no installation, no data leaving your computer.

▶ Open the Free Tool

The closing thought

The chart of UK GDP growth rates tells you the economy fluctuates. The Bootstrap CUSUM chart of cumulative UK GDP tells you the economy has been slowing for 60 years — continuously, structurally, and regardless of who was in power.

The difference is not in the data. It is in the question you ask of it.