The debate over updating data sparked reactions, with some Players going public to minimize the fact that:
Not everyone The 2022 Census data have been released so far
When, in practice, we are talking about household income information, which are among the most important for the composition of geo-demographic bases.
Estimating this type of information, in a context in which the most recent version available refers to 2010, raises a series of questions about how the work is carried out.
And unlike our colleagues, let's not claim that we have a team “of professionals who create robust methodologies”, as if that were enough for you to trust the projections.
Especially because we don't have such a big team, nor do we believe that we need that.
Understand
The preparation of our databases follows validation and consistency steps, but they are not always the same steps.
The starting point changes everything, but the master template has to be a Census, which is a universal survey, that is, it does not use samples, its purpose is to capture a portrait of the whole at a given moment.
When we have a Census, even at a significant time distance, we can use it as a good basis for projections.
After the Census, we started other smaller surveys, carried out on statistically representative samples, which are therefore more recurrent and current, such as the PNAD (National Household Sample Survey), POF (Family Budget Survey), Population Counts and others.
Scenario in April 2025
The IBGE released the main body of information from the 2022 Census in November 2024, but still without household income information.
This partial disclosure created a dilemma:
- Maintain the 2010 Census master template until income information was released
- Migrate to the 2022 Census template and estimate missing information
This type of estimate is a huge amount of work, which, oddly enough, only Mapfry mentioned having carried out.
Then the question arose, whether the other companies actually did a job of this size, something that would be a competitive advantage over the others, Why didn't they say anything about that?
The secret formula
We talk a lot about projections, visions of the future, or approximations of data from the past to the present.
But demography can also be an archaeological science, capable of investigating events that happened a long time ago, or even in imaginary worlds.
This is the size of its power, when exercised with technical mastery, it is capable of transforming statistical fragments into cohesive information.
In order to demonstrate this, we will present two powerful studies: “English Familial Demography c. 1300: A Reconstruction” and “How Many Hobbits?” , the demographic analysis of Middle Earth, the scene of the fantasy book series The Lord of the Rings.
English Familial Demography c. 1300
For their study of the population of England around the year 1300, Hugo La Poutré and Richard Paping faced a delicate challenge: the period before the Black Death was always a statistical shadow zone.
Few sources, sparse records, and yet, they manage to reconstruct a plausible picture of medieval England at its peak of population.
How did they do that?
Starting from the fragments of reliable information and creating small simulation models of the others.
The most consistent information came from parishes, which kept a record of marriages and births.
- Average age at marriage (24 years for women, 27 for men)
- Remaining life expectancy at marriage
- Average household size (5.8 people)
- Marital fecundity
These fragments are combined in an internal coherence model: if A implies B, and B implies C, then it is possible to infer D.
The result is a statistically believable world, where families take shape, lives unfold, and social structures emerge.
Middle Earth as a case study
Now imagine applying that same logic to a world that never existed.
That's exactly what demographer Lyman Stone does in “How Many Hobbits?”.
The question seems like a joke, but it's a serious population modeling exercise:
What if Middle Earth were inhabited under the same parameters of agricultural density and demographic patterns as in Medieval Europe?
Stone segments the fictional map into recognizable regions, estimates the Earth's productive capacity, and adjusts densities based on geography, climate, and history.
What emerges is surprising: a coherent portrait of a semi-real society.
Middle Earth, according to his analysis, would have about 6.8 million inhabitants, much more than preliminary estimates suggested.
That's because he doesn't start from the armies that appear in the books, but from Demographic capacity of the territory, using real historical and ecological analogies.
Demographic picture of Middle Earth
What connects England in 1300, Middle Earth and Brazil in 2025?
Everything!
Because England in the 1300s, Tolkien's Middle-Earth, and today's Brazil share the same challenge: understanding what's missing.
And this is crucial at this time in 2025, when we are still waiting for household income data from the 2022 Census.
Our ability to fill this statistical gap allows our customer reviews to continue.
And that's the same lesson we learned from the works of Stone, La Poutré & Paping, we don't need to be paralyzed by the absence of “official” data.
We can, provided that, with technical rigor and good demographic sense, we can reconstruct probable realities based on known parameters.
That's how we always did it to project and estimate populations at levels as granular as the block, but we had never gone as far as now, when estimating such an important part of a Census.
Opening the toolbox
As said, we don't have a team with many specialists, but our specialists master the many necessary specialties, such as history, mathematics, statistics, economics, and have a good deal of experience and common sense.
So let's be careful not to reveal too much, to the point of helping our competitors improve their estimates.
You will notice that the most complex models are, after all, just refined versions of a very simple question: how many people live in a place, and how do they live?
The basic population equation
The most fundamental formula of demography is this:
Future population = Current population + (Births and Immigrations) - (Deaths and Emigrations)
Every population change is the result of three factors:
- Birth rate: how many people are born
- Mortality: how many people die
- Migration: how many enter and leave the territory
Modern demography uses the Total Fertility Rate, which represents the average number of children per woman during her reproductive life.
Other methods
A source as consistent and reliable as parishes and wills are armies.
Historically, for every soldier in the field there were between 60 to 120 civilians supporting economically, logistically, and politically.
Without sources
Demography can also be measured without records, provided that the logical and physical constraints.
And with them, you can build a whole world from nothing.
Every race, whether humans, hobbits, or elves, needs land, water, calories, and time to gestate and raise children.
There are limits to the size of armies, to agricultural density, to life expectancy. These limits are known.
Lyman Stone, in the analysis of Middle Earth, uses geographically based demographic capacity models.
Given a given area and its ecological, political, and geographical characteristics, it is possible to estimate how many people it can support with pre-industrial technology.
Estimated population density table by type of territory
Coherence Rules
These methodologies do not operate in isolation, the secret lies in the coherence between the methods.
When a number calculated by density does not match the number inferred by the army, or with the number implied in family structures, something is wrong.
Demography is a system of interdependent equations.
It requires that:
- Do birth rates support the projected population
- Death rates do not contradict presumed life expectancy
- The number of soldiers is compatible with the number of men of combat age
- The average family size matches the number of houses
It is this tangle of intersections that allows us to build models of possible worlds, whether medieval, imaginary, or future.
And here we come to the practical point
By the end of April, the geo-demographic bases of the other platforms are, to a certain extent, Blindly.
But that drama is about to end, and the IBGE scheduled the release of the first part of the income information for the end of April.
In a short time, all Geomarketing platforms will have access to the essential variables for their Consumer Potential, Social Classes and other market estimates data.
But this episode is evidence of those who knew how to apply demography techniques to estimate with good precision and who could only wait.
This makes all the difference in small everyday projections, which are more difficult to compare.
Because, in the end, it's not about having a large multidisciplinary team, not least because demography goes beyond data manipulation, it is necessary to have sensitivity and common sense.
Bibliography on historical demography and population modeling
Wrigley, E.A., & Schofield, R.S. (1981). The Population History of England 1541—1871: A Reconstruction. Cambridge University Press.
La Poutré, H., & Paping, R. (2023). English Familial Demography c. 1300: A Reconstruction.
Hajnal, J. (1965). European Marriage Patterns in Perspective. In Population in History: Essays in Historical Demography (D. V. Glass & D. E. C. Eversley, eds.).
Reher, D.S. (1998). Family Ties in Western Europe: Persistent Contrasts. Population and Development Review, 24 (2), 203—234.