Opening
The Arithmetic of Belonging
In September 2026, the U.S. Census Bureau proposed the most significant change to its methodology in a generation — dropping the short form’s race and ethnicity questions, and for the specific purpose of congressional apportionment, no longer counting certain noncitizens at all.
The particulars will likely be revised, litigated, or forgotten by the time anyone rereads this. What will not change is the older question underneath them. Article I, Section 2 of the Constitution has required an “actual enumeration” of the population every ten years since 1790 — not a headcount for its own sake, but the raw material from which political power gets allocated, state by state, seat by seat. Every argument over who gets counted is really an argument over what a number is allowed to convert into.
Supporters call the proposed change a simplification. Critics call it a redefinition of whose presence counts toward a state’s power. Both, without quite saying so, are treating population as a currency that political representation is priced in.
A census looks like arithmetic. It behaves like theology.
Three thousand years before any congress argued over an enumeration clause, a king in Jerusalem ordered his own count, and the chapter that records it has never fully explained what he did wrong.
History
The Eye of the Sovereign
Three rulers, three centuries apart, each ordered a count. What connects them is not their subject but their instinct: that a number, once produced, would tell a ruler something he needed to know before he could act. David counted fighting men, to measure military capacity. William counted land and taxable wealth, to measure revenue and control over a kingdom he had just conquered. A modern census counts persons by residence, to measure political representation and the allocation of resources. Three different objects, three different purposes — and, underneath all three, the same reflex toward a number.
In December 1085, William the Conqueror sent commissioners into every English shire to record exactly what nineteen years of conquest had produced — every manor, every plow team, every parcel of taxable land. Two Scandinavian kings still coveted his crown, and William needed to know precisely what he could tax and defend before he could hold either. The survey moved fast, under a year, asking the same three questions of every holding: what it had been worth before 1066, what it was worth under its new Norman lord, and what it was worth now.1 People at the time did not call it a survey. They called it Domesday, because its verdict, once entered, admitted no appeal. The count was armor dressed as administration.
The political scientist James Scott gave this instinct a name — legibility: a state’s capacity to render a population visible enough, in categorized and countable form, to govern it at all (Scott, 1998).2 Censuses, land surveys, and cadastral maps are not incidental to statecraft in Scott’s account. They are its foundation. A ruler who cannot see his people cannot tax, conscript, allocate, or represent them.
Legibility explains why the counting happens in every case above. It does not explain what a sovereign does next — the moment, common to a Norman king, a Judean king, and a modern census bureau alike, when visibility quietly gets mistaken for entitlement. To see a people clearly is not yet to own them.
Scripture
The Census the Text Refuses to Explain
David’s census closes out the book of 2 Samuel, and the text is almost aggressively uninterested in explaining itself.
“And again the anger of the LORD was kindled against Israel, and he moved David against them to say, Go, number Israel and Judah.”
2 Samuel 24:1, KJV
Joab, David’s own general, resists before the count even begins. David proceeds anyway, and once the number is in his hands, something in him recoils — the text says only that he had sinned greatly in what he had done (2 Samuel 24:10).
The chapter never names the sin outright, and the ambiguity runs deeper than a missing explanation. It is the LORD’s anger that moves David to count, in this account; in the parallel record of 1 Chronicles 21:1, it is Satan. Scripture leaves both statements standing side by side rather than resolving them — divine protection and human self-assertion, held in the same sentence without a seam between them. Underneath that tension sits one plainer question the text keeps circling: what did David actually want the number to prove?
The clearest answer available comes not from 2 Samuel alone but from a contrast Scripture itself sets up. Numbers 1 opens with God commanding Moses to count Israel’s fighting men, tribe by tribe, ahead of the march into the wilderness — a census taken at God’s initiative, for God’s purpose. What numbering the people had always required, in Exodus, was a ransom — a small payment that stood in for a larger claim: that Israel’s lives belonged to God, and a census could only be taken in a way that acknowledged this ownership rather than quietly asserting a different one (Exodus 30:12). David’s count shares the same form as the one in Numbers and skips everything that made it holy — no command, no ransom, no acknowledgment. A king does not merely miscount when he takes that shortcut. He counts as though the people before him need no acknowledgment at all — the same taking-without-asking this blog has traced before in David reaching for what wasn’t his to reach for (David and Bathsheba).
What did David want the number to mean? The text lets that question outlive him.
Pattern
Limits of the Ledger
Counting is not, in itself, a violation. More often it is the opposite. A person left off a census can also be a person left out of representation, funding, or protection — being counted is frequently how a person becomes visible enough to be governed justly rather than simply governed. That is the strongest case for an accurate census under any methodology, and it deserves to be said plainly, because nothing here argues that being counted is dangerous.
The danger sits somewhere narrower: not that a government counts its people, but what it believes the resulting number then gives it permission to decide. A ledger can record how many people live inside a state’s borders. It cannot record what any one of them is worth, or who a person actually is when no one is checking the register — the same gap this blog has examined in what integrity looks like with no one watching (Honesty When No One Is Watching).
That fuller question is not one this piece can settle for the American proposal now working through public comment, and it should not try. Serious, non-cynical arguments exist on more than one side of how citizenship, residence, and apportionment ought to be defined — legal, constitutional, administrative arguments, made in good faith by people who disagree. The older question underneath all of them survives any particular answer a legislature reaches: whichever formula a nation adopts, what does it imagine the resulting number entitles it to decide about the people inside it?
Scott’s legibility answers half of that question — a ruler has to see a population before he can govern it. Scripture answers the other half, in the negative: making a people visible does not make them yours.
Closing
The Crown and the Uncounted
David’s census produced an exact figure — a kingdom’s strength reduced to a tally small enough to hold in one hand. William’s commissioners produced a book so thorough it took the name of a day of judgment nothing could appeal. A modern census will produce, in time, a figure that decides how many seats a state holds in Congress for a decade.
Three counts, three purposes, one temptation running underneath all three: to let a number answer a question it was never built to answer.
The temptation David faced was never really about arithmetic. It was about what a number is allowed to prove.
A census can tell a government how many. It cannot tell a ruler whom they belong to.
In the previous installment, we examined what it means for a betrayed man to test whether the brothers standing before him had actually changed, in “The Reveal.”
1. William the Conqueror’s 1086 land survey, compiled from county-by-county inquests, remains among the earliest and most detailed surviving population-and-property surveys of medieval England (Britannica, 2026; ONS, n.d.).
2. James C. Scott, Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed (New Haven: Yale University Press, 1998).
New essays every week. Delivered quietly.