Most of England's most deprived neighbourhoods have stayed that way since 2019

Why there is no IMD2019-vs-IMD2025 trend chart

Given the indicator, geography and domain-composition changes above, a chart plotting IMD2019 and IMD2025 scores or ranks for the same LSOA on one time axis would conflate genuine deprivation change with definitional/methodology change, in a way readers cannot disentangle from the chart alone. The one figure MHCLG itself publishes bridging the two editions — 82% decile persistence — is shown here as a headline continuity statistic, not as license to build a two-point trend line from it.

• Indicator basket materially expanded and changed: 55 indicators in IoD2025 versus 39 in IoD2019 — 20 are entirely new, 14 were "updated and significantly modified", and only 21 were "updated without significant modifications" (Technical Report §3.3.4).
• LSOA geography changed: IMD2015 and IMD2019 used 2011-Census LSOA boundaries; IMD2025 uses 2021-Census LSOA boundaries. MHCLG states "the boundaries of approximately 6% of the 2011 LSOAs were modified" (Technical Report §1.2.2, footnote 2) — so even an LSOA whose code appears unchanged may not cover the same footprint.
• Crime domain rebuilt: 4 indicators in IMD2019 (violence, burglary, theft, criminal damage) became 8 in IMD2025 — three separate violence-related indicators, a new public order/weapons-possession indicator, and a new anti-social-behaviour indicator — drawing on newly-granted Home Office Data Hub access and a fuller set of notifiable-offence categories, over 6 years of crime data instead of IMD2019’s 2 (Technical Report §4.7.24–§4.7.27).
• Living Environment domain indicators changed: the IMD2019 "households lacking central heating" indicator was dropped as no longer discriminating (too few households now lack central heating); 3 new indicators were added — EPC-based housing quality, lack of private outdoor space, and road/rail/aircraft noise pollution; the indoor/outdoor sub-domain weighting shifted from 67%/33% (IMD2019) to 70%/30% (IMD2025) (Technical Report §4.9.38–§4.9.41).
• Income domain indicator basis shifted with the ongoing national rollout of Universal Credit, only partially reflected in IMD2019’s methodology and still "ongoing development work" as of IMD2025 (Technical Report §4.3.3).
• Domain weights themselves are UNCHANGED between IMD2019 and IMD2025 (Income 22.5%, Employment 22.5%, Education 13.5%, Health 13.5%, Crime 9.3%, Barriers to Housing and Services 9.3%, Living Environment 9.3%) — the discontinuity is in what each domain measures internally, not in how the 7 domains combine.
• The IMD is, by MHCLG’s own description, "a relative measure of multiple deprivation" (Technical Report §3.1.3) — every score/rank is relative to every other LSOA at that one point in time, not an absolute scale. Even a methodologically-identical re-run would not support "this LSOA’s score fell by N points" as a real-terms improvement claim, because a change in relative rank always reflects both that area’s own change and every other area’s change together.