Mexico stopped using CONAPO population to compute its national mortality rate in 2020, and its statistics institute says so in the chart footnotes
Translated from the Spanish original.
On November 10, 2026, INEGI (the national statistics institute) published press release 105/26 with the definitive 2025 Registered Deaths Statistics: 792,187 deaths registered, of which 771,178 (97.4%) occurred in 2025 and 21,009 in prior years; a decrease of 27,485 cases compared with 2024; and a crude rate of 606 deaths per 100,000 inhabitants, 24 units below the previous year. The footnote to chart 1, which plots the 2016-2025 series as a single line, warns that the denominator is not the same throughout the series: "The crude rate of registered deaths per 100,000 inhabitants is presented as a reference to facilitate comparison between years. The denominator for computing the rate for the 2016 to 2019 period was generated from the 1950 to 2019 Demographic Reconciliation of the Consejo Nacional de Población. The denominator for the years 2020 to 2025 corresponds to the population estimate prepared by INEGI on the basis of the Housing Sampling Frame. The population was computed with reference to June 30 of the corresponding year". The same warning reappears at the foot of charts 2 and 3. That is, since 2020 Mexico's official mortality rate has been computed on a population estimated by INEGI itself rather than on CONAPO's projection (CONAPO is the national population council), which is the one the Health Ministry distributes and the one underpinning sectoral planning. The two rates in the same release allow that denominator to be estimated: 792,187 divided by 606 per hundred thousand implies 130,723,927 inhabitants, and 771,178 divided by the rate of 590 for deaths registered and occurred implies 130,708,136 —both own calculations, subject to the rounding of the rates to whole units. CONAPO's standing projection put the mid-2025 population at 133.4 million.
Strategic interpretation
Systemic: Mexico has two official populations and the difference is around 2.7 million people, roughly 2% of the country —an own calculation from INEGI's implied denominator and CONAPO's reported projection, not a figure published by anyone. The two poles: INEGI, which measures and since 2020 prefers its own Housing Sampling Frame estimate, and CONAPO, which projects and whose output remains the reference population for sectoral planning. The break was not hidden —it sits at the foot of three charts, where INEGI notes the rate "is presented as a reference"— but no release was located explaining why the institute that publishes the statistic stopped using the projection of the council that projects it. INEGI gains precision; the country loses a single denominator and, with it, the comparability of every per capita indicator. Consequence: It lands on anyone computing per hundred thousand inhabitants: Health Ministry targets, federal transfers, SESNSP crime rates, education coverage. From that it follows that two areas of the same government can report different rates for the same phenomenon without either lying. And from that it follows that the direction of the bias matters: if the real population is smaller than projected, declared coverage is understated and rates inflated; if larger, the reverse. No one has published which it is. Over 3-10 years the arbiter is the 2030 Census: there it will be known which population was closer, and the answer will reframe a decade of indicators. Worldview: That the denominator is a technical detail and not a policy decision — that "the population of Mexico" is a fact one looks up and not the product of a method someone chose. What to watch: Whether CONAPO publishes a new reconciliation incorporating the Housing Sampling Frame or whether the two series keep running in parallel; whether any agency states which denominator it uses; whether INEGI explains the change in a methodological document; and the 2030 Census as arbiter.
INEGI, press release 105/26, Registered Deaths Statistics (EDR), November 10, 2026. The PDF was downloaded and read page by page in the main thread on November 19, 2026: the figures and the footnote to chart 1 are transcribed from the document. Methodological warning from this run: the first automated extraction of the same PDF returned 755,529 and 748,089 deaths, figures that do not correspond to the release; they were discarded after direct reading. The implied denominators are own calculations. CONAPO's projection of 133.4 million for mid-2025 comes from press coverage of its 2020-2070 Population Projections and is declared as reported, not verified against CONAPO's database. Independent verification of November 19, 2026: it read the PDF on its own, confirmed every figure in the release, restored the full footnote this record had truncated at both ends, and corrected the implied denominator, which had been divided wrong: 130,723,927, not 130,724,587.
Chapter signal radar
Published on November 19, 2026