Comparing two periods
3 min read
Lesson objectives
- Create a bridge between two population series when boundaries, residence rules or editions change.
Please sign in to use all the tools
Once you sign in, the reading tools are yours: larger or smaller text, typeface, line spacing, focus mode, highlighting and bold, and notes saved to your account and gathered in My study.
Comparing two periods
A count difference between years can mix actual population change, changes in the geographic or residence definition, and revisions to older values in a newer source edition. Separate these layers before interpretation. Start a table of reference year, geographic boundary, residence rule, source and edition. If the boundary changed, seek a retabulation of the earlier year to the later boundary. If no bridge exists, give a limitation rather than a precise change rate. The same district name does not guarantee the same footprint; an administrative change may transfer whole blocks.
A producer may revise an earlier year's estimate after receiving a census or register update. Use a consistent edition for comparison where possible. Do not contrast an old edition's earlier value with a new edition's later value as if the entire gap were demographic. Preserve the original and revised values in the working record and explain which one is used. If the analysis calls for midyear population, obtain a documented midyear estimate rather than silently substituting an end-year count. For a simple percentage change, compute (new-old)/old×100 only after confirming the same population definition. A multi-year percentage is not automatically an annual compound growth rate.
Write the interpretation from what the evidence supports: ‘The comparably defined count rose from X to Y between these dates.’ Even that net change does not identify its cause. Do not attribute it to migration or births without reliable component evidence. If no boundary reconciliation is available, state that direct comparison is invalid and specify what would repair it. That is a useful analytic conclusion, not a failure to calculate.
For an interval of two or more years, do not divide total percentage change by the number of years and call the quotient a compound annual rate. When appropriate, the compound annual rate is ((new/old)^(1/years)-1)×100. A rise from 100 to 121 over two years is 21% in total and 10% compounded annually because 100×1.1×1.1=121. The formula cannot rescue a zero baseline or inconsistent definitions and boundaries. Repair comparability before annualizing.
Lesson summary
Align boundary, inclusion rule and edition before interpreting count change, and annualize a multiyear change only with a stated method.