How sustainability targets describe baseline and coverage
A sustainability target is a comparison rule attached to a defined population and period. Reading only its percentage can lose the baseline exception, the reporting window or the meaning of success. This study examines two historical public disclosures, then tests those reading decisions with a wholly fictional target register.
The cases are selected UK departmental reports, not a representative sample or a ranking. They examine the completed 2021–25 Greening Government Commitments cycle; this article does not describe current programme requirements.
DESNZ A baseline exception below the summary
The DESNZ 2024–25 performance report describes a 2017–18 baseline and coverage of its core department plus six public bodies. Its summary lists a 30% domestic-flight emissions reduction target and a reported 45% reduction. The following narrative explains that incomplete 2017–18 flight data led to use of 2018–19 as the flight baseline.
Our interpretation: the general column heading cannot safely populate every target’s baseline field. Preserve the flight-specific exception with that row. Domestic-flight emissions also remain a narrower activity than the report’s overall emissions category; a broad travel label would lose that distinction.
Defra A reporting window and different target forms
Defra’s performance disclosure, updated 10 April 2024, describes progress for 1 January–31 December 2023 against 2017–18, with targets due by March 2025. It distinguishes quantity reductions from final-state ratios: paper use has a 50% reduction target, while recycling must exceed 70% and landfill must be below 5%.
Our interpretation: the update date, the twelve-month measurement window and the target deadline belong in separate fields. A percentage reduction from a baseline is a different rule from a percentage of current waste. Preserve the strict above/below wording instead of converting every percentage into a reduction target.
What can be compared between the cases
Both disclosures reward reading beyond the headline. One needs a row-specific baseline exception; the other needs a separate reporting window and comparison operator. They support studying target design, not pooling departmental results.
For an evidence review, create one record per target statement. Keep the exact source edition, named reporting organisation, included activities, reference period, target period, metric unit and comparison rule together. If a source gives only a year, do not silently expand it into dates. If the relevant organisational perimeter is unresolved, record that limitation before calculating progress.
Treat contextual notes as part of the target’s identity. A target on purchased electricity, one on all building energy and one on a wider corporate footprint may overlap without being interchangeable. An organisation may meet a broad total while missing a narrower component target. A source-selection spreadsheet should preserve those relationships rather than force a single traffic-light verdict.
Original case Four tests for one fictional organisation
Everything in this case is invented, including the targets. A fictional reporting owner approves four tests for a complete twelve-month reporting period ending 31 March 2026. All supporting values are exact for this exercise; activity coverage and methods are assumed comparable to the stated baselines. The example illustrates interpretation, not a recommendation for target ambition.
Overall emissions: reduce by at least 30% against 2018–19. Baseline 1,000 tCO2e; current 640 tCO2e.
Domestic-flight emissions: reduce by at least 40% against 2019–20. Baseline 100 tCO2e; current 72 tCO2e.
Recycling: recycled material must exceed 75% of current waste. Recycled 150 tonnes; total waste 200 tonnes.
Landfill: landfilled material must be below 4% of current waste. Landfilled 7.98 tonnes; total waste 200 tonnes.
Domestic flights are included within overall emissions for each relevant period. Recycling and landfill are disjoint destinations within the same 200-tonne waste total. The remaining 42.02 tonnes have other destinations, which are outside these two tests. These relationships are stipulated inputs; a real review would need evidence for them.
Overall emissions have fallen by (1,000 − 640) ÷ 1,000 = 36%. The 30% reduction requires a result no higher than 700 tCO2e, so the current figure meets this test by 60 tonnes. Flight emissions have fallen by (100 − 72) ÷ 100 = 28%. Their separate 40% target requires no more than 60 tCO2e; the current figure remains 12 tonnes above it.
There is no contradiction in these two decisions. They use different reference periods and different activity coverage. Do not add the two targets’ emissions ceilings: the flight amount is a component of the broader footprint, not an additional emissions category. Likewise, the broader target’s success does not close the flight-specific shortfall.
If a summary row copied the overall baseline year into the flight target, the record would already be wrong before any arithmetic began. The correct historical flight quantity is tied to its own approved reference period. An analyst cannot choose whichever past year produces a favourable percentage or substitute a neighbouring total just because both are expressed in tCO2e.
Apply the operator before rounding the result
The fictional recycling result is exactly 150 ÷ 200 × 100 = 75%. It does not meet a requirement to exceed 75%. If the approved wording were instead “at least 75%”, the same evidence would pass. Changing that operator changes the commitment; it is not an editorial simplification for the reporting team to make.
Landfill is 7.98 ÷ 200 × 100 = 3.99%, which meets the requirement to remain below 4%. Displayed to one decimal place, it becomes 4.0%. Decide against the unrounded value under the stipulated rule, then show sufficient precision or a note explaining the result. A displayed value alone should not overturn the underlying calculation.
These conclusions depend on exact inputs and the stated comparison rules. In a real dataset, measurement uncertainty, estimates or rounding of the underlying tonnages could matter near a threshold. Record the actual precision and the applicable evaluation policy. Do not claim a robust margin from extra decimal places generated by a spreadsheet.
Retain the decision without overstating the outcome
The fictional review produces four separate outcomes: overall emissions meets its test; domestic flights does not; recycling does not; landfill does. This is not a 50% sustainability-performance score. The targets differ in subject, scale and overlap, and no weighting or aggregation rule has been supplied. A count of passed tests would answer a different question.
Retain each calculation with its baseline version, measurement window, activity boundary, exact operator and evidence reference. Use a separate issue field for missing definitions. When information changes, revisit the affected target rather than silently refreshing all rows with one organisation-wide baseline date.
The emissions-change guide explains movements between inventory periods. This study addresses an earlier reading decision: identifying the exact commitment against which a result is being tested. Passing a target’s arithmetic test does not itself establish an intervention’s causal effect, the adequacy of the target or independent assurance.
Sources checked 14 October 2026. Contains public sector information licensed under the Open Government Licence v3.0. Analysis and fictional experiments are Nitium’s own; no source table or diagram is reproduced.
Comments