The Sum of the People

How the Census Has Shaped Nations, from the Ancient World to the Modern Age


By Andrew Whitby

Formats and Prices




$38.00 CAD

This item is a preorder. Your payment method will be charged immediately, and the product is expected to ship on or around March 31, 2020. This date is subject to change due to shipping delays beyond our control.

This fascinating three-thousand-year history of the census traces the making of the modern survey and explores its political power in the age of big data and surveillance.

In April 2020, the United States will embark on what has been called “the largest peacetime mobilization in American history”: the decennial population census. It is part of a tradition of counting people that goes back at least three millennia and now spans the globe.

In The Sum of the People, data scientist Andrew Whitby traces the remarkable history of the census, from ancient China and the Roman Empire, through revolutionary America and Nazi-occupied Europe, to the steps of the Supreme Court. Marvels of democracy, instruments of exclusion, and, at worst, tools of tyranny and genocide, censuses have always profoundly shaped the societies we’ve built. Today, as we struggle to resist the creep of mass surveillance, the traditional census — direct and transparent — may offer the seeds of an alternative.


And Moses and Eleazar the priest spake with them in the plains of Moab by Jordan near Jericho, saying, Take the sum of the people, from twenty years old and upward; as the Lord commanded Moses and the children of Israel, which went forth out of the land of Egypt.

—Numbers 26:3–4, King James Version

According to the Bible, Mary and Joseph traveled from Nazareth to Bethlehem to register for a Roman census. The Nativity Trail, a modern tourist route, loosely retraces the biblical journey across a complex political landscape. (Map data credit: Natural Earth, UN OCHA oPT, Imbach Reisen.)

In the central part of the West Bank, in the vicinity of Jerusalem, the official frontiers are especially convoluted. Walking through here would be difficult, so instead our route veers southeast, toward the Jordan Valley and Jericho. The Norwegians’ guide, Nedal, explains that Mary and Joseph may have taken a similarly circuitous route two thousand years ago, as they answered the call of the Roman census. An ancient road known as the Way of the Patriarchs connected Nazareth and Bethlehem directly, but it passed through the territory of the Samaritans, a group related to, but sometimes hostile to, the Jews. There’s no clear consensus on this; other evidence suggests that Jewish travel through Samaritan lands was routine. But even this best case meant a journey of around eighty miles. The terrain in this region is beautiful but often stony and unforgiving. At times Ahmed has to cajole Casimiro, the donkey, into continuing. It would be an unpleasant trip, I imagine, for somebody in the final days of pregnancy—infuriating, even, given that the purpose of the census was Roman tax collection.10

Modern censuses are not used for collecting taxes. Nor do they require such arduous journeys of those they count. Each of the Palestinian Authority’s three censuses of the West Bank and Gaza—1997, 2007, and 2017—was conducted in accordance with the UN recommendations. Two thousand years ago, Roman census records would have been recorded onto papyrus scrolls at central locations. Today, Palestinian enumerators travel to the homes of the people they are tasked to count, armed with paper forms and tablet computers. In richer countries, people often enumerate themselves, receiving and returning a census form by mail or completing it online. In a handful of countries, censuses are now virtual, compiled from a register of the entire population that is kept continuously up to date.

Palestine is not there yet. Even in 2017, its tablet-toting census takers followed a fairly traditional process, first mapping out the areas where people were thought to live and then traveling door-to-door, collecting information about each person in every household they encountered. It was the enumerators, and not those being enumerated, who had to contend with the complex topology of the Palestinian proto-state: the manned and unmanned checkpoints, the walls, gates, ditches, and barbed wire. Census taking everywhere involves more logistics than statistics, but in Palestine it involves diplomacy as well.

I meet the census team from the Palestinian Central Bureau of Statistics (PCBS) at their office in Ramallah, the administrative capital of the Palestinian Authority. Ramallah is located to the north of Jerusalem, just off Highway 60, the modern-day equivalent of the Way of the Patriarchs. Mary and Joseph might have passed here on the way to Bethlehem, which is only sixteen miles further south. Today that distance seems much greater, stretched by the tangled frontiers that surround Jerusalem. Since our nativity route took us west to avoid all that, I have circled back afterward, on my own.

Once I find the right building, I’m joined by four senior officials from the bureau, a man and three women, some of whom have been involved in Palestinian census taking since its modern reintroduction in 1997. Tea is brought, and we start to discuss the complexities of conducting a census in Palestine.11

The greatest challenge for each of the three censuses has been East Jerusalem. After the 1948 war the city was split: Israel in the west and Jordan in the east (a UN plan for an international city came to nothing). After Jordan’s defeat in the 1967 war, the eastern part came under Israeli control. At that time, a census was taken by Israel, and Palestinians present were given a status of permanent residence in Israel. Today they are issued with Israeli identity cards and can in theory apply for full Israeli citizenship. Most—reportedly 95 percent—have not. Their status remains somewhat precarious.

The Palestinian Authority views East Jerusalem as occupied Palestinian territory, an integral part of the future Palestinian state; the people living there, then, should naturally be included in the census of Palestine. At the first census in 1997, the Authority set out to do exactly that. It was an act of open defiance against Israel, a continuation of Palestinian resistance by statistical means. The then-head of Palestinian statistics called it “a civil intifada,” borrowing the word given to the period of unrest that had been brought to an end with the signing of the Oslo accords. Israel argued that the Authority’s actions violated the accords, moving to outlaw Palestinian census taking in East Jerusalem. There was at least one arrest, and the census of East Jerusalem was halted.12

In 2017, seventeen arrests were reported of people associated with the Fatah Party, which rules the Palestinian Authority. Israeli police accused them of “taking part in activity related to a population census.” At the time, the Palestinian Central Bureau of Statistics denied operating in East Jerusalem and claimed the arrests “had nothing to do with the population census.” And yet there in the final census report are census counts for East Jerusalem.13

I ask the Palestinian statisticians about this. Their response is one of calculated vagueness, echoing the official denials. (I infer that the enumeration was undertaken by Fatah affiliates already living in Jerusalem, rather than employees of PCBS.) While enumerators used electronic tablets for the rest of the West Bank and Gaza, paper forms were used in East Jerusalem. The questionnaire was shorter than elsewhere in Palestine. These strategies seem designed to avoid drawing too much Israeli attention. Needless to say, this is not a situation addressed in the 299 pages of the UN’s Principles and Recommendations.

Two other major logistical difficulties confronted the 2017 census takers. The first was the Separation Barrier, which did not exist in 1997 and had grown substantially since 2007. Travel to and from some Palestinian communities is restricted to people who are registered as living there, which makes staffing a census harder. “Mostly we recruit people from the same governorate, from the same locality, especially for localities behind the separation wall,” they tell me. The electronic tablets allowed data to be transmitted wirelessly to the head office in Ramallah, reducing the amount of paper shuffling required and eliminating the possibility of seizure by Israeli soldiers. A similar situation held with respect to Gaza: travel back and forth was minimized by hiring local staff and using videoconferencing and other technology. Even with such measures, the 2017 census required, they observe, “emergency plans, all the time.”

When the interview is finished, I take the number 218 public bus back to Jerusalem. It passes through the famous Qalandiya checkpoint, which divides Jerusalem from the West Bank. While we wait at the checkpoint, I skim through the official Palestinian census report once again. I make note of the population totals: 2.9 million in the West Bank, including 281,163 furtively enumerated in East Jerusalem; 1.9 million in Gaza; 4.8 million altogether. In many respects it is a typically dry, technical document, replete with obligatory discussions of statistical arcana such as nonsampling error, coverage percentages, and Whipple’s index.14

But politics is never far away. The census is a “pillar of state building” and “a genuine expression of national sovereignty.” Enumeration was impeded by “the procedures of the Israeli occupation and obstacles including the Annexation Wall and settlement expansion.” That wall, which looms over the bus as we leave the checkpoint and reenter Jerusalem, “suffocates those living behind it.” Even the usually soporific methodology section notes “instability resulting from continuing Israeli aggression, confiscation of land and isolation of the population in Palestinian Localities.”15

While I do not doubt the statistical rigor of the Palestinian census, its parallel geopolitical purpose is unmistakable. Even the now-predictable arrests play into that: there is no imminent prospect that the Palestinian Authority will actually govern East Jerusalem, so—official denials notwithstanding—its purpose in counting people there seems less administrative than symbolic: to publicly assert its territorial claim. As we pull into the bus station near the ancient Damascus Gate of the old city of Jerusalem, I note, before closing the report, one more potent symbol: fieldwork was completed on December 24, 2017—Christmas Eve.16

Census taking makes sense as a tactic of nation building. The conventional definition of a state under international law requires a permanent population, a defined territory, a government, and the capacity to enter into relations with other states. A census is surely the most direct way to demonstrate a permanent population. It’s also evidence of a functional government.

In 1993, in a critique of the first Oslo accord, the Palestinian-American scholar Edward Said called for an immediate census: “not just as a bureaucratic exercise but as the enfranchisement of Palestinians wherever they are… an act of historical and political self-realisation outside the limitations imposed by the absence of sovereignty.” Said’s qualification—or rather lack of qualification—“wherever they are” was significant. Today there are around thirteen million people who might be considered Palestinian, by virtue of having once lived in historical Palestine or being born to parents or having grandparents who had lived there. Around half live in the West Bank, Gaza, and Israel, with the balance spread between Jordan, Lebanon, Syria, and elsewhere. In holding a census, Said argued, they would “come close to constituting a nation rather than a mere collection of people.” But the census that followed four years later in 1997 did not count Palestinians wherever they were. Instead it conformed to international norms, counting people only within the defined territory the Palestinians claimed. It reinforced a particular, concrete assertion of statehood, at the expense of Said’s vision of national self-realization.17

Census takers are constantly grappling with lines of citizenship, nationality, ethnicity, and race. Citizenship is the clearest of these, while the other categories are much fuzzier, grounded in shared history, culture, and ancestry. Israel itself illustrates this complexity. In 2013, a group of activists petitioned Israel’s Supreme Court to allow them to record their nationality as “Israeli” in the state population register. The court declined, concluding that the state of Israel contains within it people of Jewish, Arab, and other nationalities, but nobody of Israeli nationality. It allowed that such a nationality may one day come to exist, but upheld the lower court’s ruling that “technical-statistical registration” was not a process that could have this effect.18

Arguably, the court was wrong as a matter of fact: history abounds with examples where nationality, race, and ethnicity—far from emerging organically—were assigned, or even created, precisely by “technical-statistical registration.” Prior to the widespread use of self-enumeration in 1960, an American’s race, on the census, was largely determined by the enumerator—who was, in all probability, a white person. That is no longer the case. Censuses in many countries increasingly treat nationality, race, and ethnicity as a matter of self-identification, refusing to arbitrate or second-guess such slippery categories. Of course, that’s not quite right either. A person cannot simply arrive in Israel and successfully self-identify as Jewish. These concepts may not be objective, but neither are they completely subjective.

While I was in Israel, a debate erupted following a statement from the prime minister, Benjamin Netanyahu, that “Israel is the nation-state of the Jewish People—and them alone.” Netanyahu was on firm legal ground, having passed a law to that effect in 2018. Israel is now a state in which the Jewish nationality is paramount; there is no Israeli nationality because, by law, the Israeli nationality is Jewish.19

That doesn’t leave much room for a viable one-state solution, but as we discussed this over long days of walking, I found my Swiss companions more optimistic. As they reminded me, national identities are not fixed. In a continent that spent much of its history bloodily rearranging its constituent states along ethnic, religious, and linguistic lines, Switzerland stands as an exception, with four official languages. Roberto, a retired English teacher, German-speaking, but with a name that recalls his family’s Italian origins, taught me a word the Swiss use to describe themselves. Willensnation: a nation brought into being, and held together, by sheer force of will. It’s a hopeful story.

The Church of the Nativity, in Bethlehem, is the endpoint of the Nativity Trail, by tradition the site of the stable where Jesus was born. To my (and nobody else’s) disappointment, Luke doesn’t elaborate on how or where the Roman census was taken, or whether the family even made it to be registered, given Mary’s presumed state of indisposition and the challenge of caring for a newborn baby. So I too ended my pilgrimage here.

It was unlike any place of worship I had ever visited, not so much a church as a set of interconnected churches—Greek Orthodox, Armenian, and Roman Catholic—all sharing one site. This arrangement, known as the Status Quo, dates back to Ottoman times. It is replicated in miniature in the holiest part of the complex, the Grotto, supposedly the exact place where Mary gave birth. This subterranean space is lit by fifteen hanging oil lamps, whose numerical configuration is specified precisely by the Status Quo: six are Greek, five Armenian, and four Roman Catholic. These proportions have no contemporary meaning that I can discover: they were simply frozen in place, in 1852, by a Turkish sultan tired of religious squabbles. As I read these facts, by the lamplight of the Grotto, I was reminded of another status quo, holding tenuously 150 miles to the north.

Lebanon is the prodigal son of global census taking. Its last full count was in 1932, under French rule. When independence came, in 1943, power was divided between various religious sects on the basis of those decade-old statistics. No individual group—Maronite Christian, Sunni, Shia, and on down a list of seventeen officially enumerated—dominated, though Christians overall just outnumbered Muslims. So Lebanon created its own status quo: top government positions were allocated in proportion to the 1932 population, and the parliament and civil service were established with an exact ratio of six Christians for every five Muslims.20

Over time, the ratios between these groups in the population began to drift, even as the 6:5 agreement remained fixed. Quantifying the change and revisiting the agreement was considered too destabilizing, so successive governments, from 1943 until 1975, chose instead to ignore it, refusing to count the people of Lebanon. At the conclusion of the civil war that lasted from 1975 to 1990, the ratio was amended to 1:1, Muslim and Christian parity, reflecting the belief—almost certainly correct, but untested by official data—that the Christian proportion had declined.21

You could, reasonably, see this new arrangement as a mere political compromise, divorced entirely from demography, just as every American state receives two senators, regardless of population. But that’s not how the people of Lebanon see it: they subscribe to the democratic ideal that political power should reflect numbers. Since in practice it cannot—at least not without the risk of violent conflict—a fresh census would be a provocation. An official, complete count of Lebanon’s estimated six million people does not seem likely any time soon.

As I exited the Church of the Nativity and stepped, blinking, into Manger Square, I wondered whether the proponents of a one-state solution for Israel and Palestine imagine that it would turn out more like Switzerland or like Lebanon.

I return from the Holy Land to the United States on April 1, 2019—exactly one year before this country’s next census day. In most decades, by this point, the main parameters of the enumeration—and certainly the questions that will be asked—have been decided. Not this time. A debate is raging over a late order from the secretary of commerce to add one more question to the census, against the near-unanimous advice of Census Bureau experts: Is this person a citizen of the United States?

As tends to happen in this litigious nation, the debate has become a lawsuit—actually, multiple lawsuits—appealed all the way to the Supreme Court. The plaintiffs argue that this last-minute addition will dissuade noncitizens, especially undocumented people, from participating in the count. Oral arguments in Dept. of Commerce v. New York are scheduled for April 23. In an editorial, the New York Times bills it the highest-profile case of the court’s term. While I’ve been looking elsewhere for the story of the census, it has found me in my adopted home.22

April 23 arrives, and I read the transcript as soon as it is released. In their attempts to interrogate the legitimacy of the proposed question, the justices grapple with historical and international comparisons. In the first minute, the solicitor-general, presenting the case for the Department of Commerce, declares that such a question “has been asked as part of the census in one form or another for nearly 200 years.” He is interrupted by Justice Sotomayor, who contends (correctly) that this is an oversimplification. Later, Justice Kavanaugh observes that many other countries ask about the citizenship of respondents and that the UN includes citizenship among its recommended census topics. “The question,” he says, is does “international practice, that UN recommendation, that historical practice in the United States, affect how we should look at the inclusion of a citizenship question in this case?”

This book’s answer is yes. A historical and international perspective is essential, not just to understand the issues that were at stake in that now-resolved case but to understand the meaning of the modern census more generally. The census did not appear fully formed in some particular time and place but developed in a slow, continuous interplay of ideas from around the world. The belief that vexes Lebanon, that each census should be followed by a redistribution of political power, is an essentially American idea. The notion that people should be enumerated individually, rather than simply as anonymous members of a household, is Scandinavian. The modern method of testing census accuracy was trialed first in India. Intellectually, the census extends beyond borders, across oceans, and back through centuries, even if individual censuses occur in one place, at one time.

In the chapters that follow, I argue that the census is not just a collection of separate national projects but a human project. My aim is to tell this larger story: to show how the institution of counting people has evolved, how it has changed as societies changed, and how it has sometimes changed those societies in turn. In doing so, I draw on a scholarly movement that, since the late twentieth century, has begun to treat the study of statistics as an object of study itself. This has produced some illuminating academic accounts of census taking in particular countries, but none that attempts, as I do here, to sketch the global history for a nonexpert reader.

Scholars have often cast the census as an instrument of power: as a mechanism of “state formation” and control, following the influential French philosopher Michel Foucault; as an attempt to make society “legible,” in the terminology of political scientist James Scott; or as a top-down project of nation building by the “classifying mind of the colonial state,” as the anthropologist Benedict Anderson put it. There’s a great deal of insight in these perspectives: even the simplest, most essential result of a census—the knowledge of a population’s size—can have a kind of power, as the Malthusianism of the nineteenth century and its twentieth-century echo illustrate.23

But the census can be a tool, too, of the powerless. It has been embraced, appropriated, and even subverted by those being counted. It has served as a medium for individual and minority self-expression. Even under the strictest regimes—Nazi occupation, for example—people have found, in enumeration, a canvas for protest. Granted, a census is never a blank canvas, but more like a paint-by-numbers in which authorities define, by setting the questions and sometimes the possible answers, both the outline and the palette. Still, by answering questions against expectation, by writing—sometimes quite literally—between the lines, or in the last resort, by absenting themselves entirely, otherwise disempowered people have conspired to reject these impositions. All this is possible because the census has a fundamentally democratizing character: it requires mass participation. It is neither wholly of the state nor of the people but exists as a continuous negotiation between them.24

Perhaps the most important phase in that negotiation, achieved in the eighteenth and nineteenth centuries, was the separation of the census from the individual obligations of taxation and military service that so often characterized it earlier. This happened at first casually, almost by accident, but eventually evolved into something closer to a promise. For all the tablet computers and wireless transmissions, this is the innovation that most fundamentally divides the Palestinian census of 2017 from the census of Quirinius on the same terrain, two thousand years earlier. Today the census is something unique: a way for the state to see the people, without seeing any individual person.

That may also be the greatest threat to its continuation, for while the traditional census has shed its role in mediating the relationships between individual citizens and their governments, those relationships have only deepened. Over the twentieth century, individual obligations were joined by individual entitlements arising from the centralized welfare state. As a result, the census today competes with many other sources of information: applications, returns, registrations, records and disclosures, each facilitating some direct relationship between the citizen and the state, each a way to render the citizen “known.”25

For now the population census still sits at the center of that constellation. But this does not guarantee its future. In a world of driver’s licenses and passports, tax returns and benefits checks, fingerprints and retina scans, hourly social media status updates and minute-by-minute location tracking, the traditional census seems increasingly anachronistic—as one group of sociologists put it, “an outdated high modernist invention.” It is infrequent, expensive, and bound by strict privacy rules.26

Some countries have now abandoned the decennial enumeration altogether. Instead, they maintain population registers, databases of their citizens and visitors that are kept continually up-to-date, so accurate and current as to render a special, once-a-decade enumeration superfluous. This started in the Nordic world and is now spreading to other countries in Europe and beyond. It’s very likely that population registers represent the next phase in the long history of counting and classifying people.

For now, though, the traditional census still rules, and it is the heart of this book. There is something special about an actual enumeration, with its proverbial army of canvassers. In 1882, Leo Tolstoy enlisted in one such army, going door to door in Moscow’s Khamovnitcheskiy quarter to count its residents. He took his job seriously, coming face to face with every type of working-class Muscovite: “master-artisans, bootmakers, brush-makers, cabinet-makers, turners, shoemakers, tailors, blacksmiths… cab-drivers, young women living alone… female pedlers, laundresses, old-clothes dealers, money-lenders, day-laborers, and people without any definite employment.” The census, he wrote, “furnishes… a mirror into which, willy nilly, the whole community, and each one of us, gaze.”27

Far from a dry statistical exercise, the census is ultimately about people; it is a form of quantitative social history. In 1867, then congressman James Garfield noted that

till recently the historian studied nations in the aggregate, and gave us only the story of princes, dynasties, sieges and battles. Of the people themselves—the great social body with life, growth, sources, elements, and laws of its own—he told us nothing. Now statistical inquiry leads him into the hovels, homes, workshops, mines, fields, prisons, hospitals, and all places where human nature displays its weakness and its strength.28

This is the story, too, of people: those who were counted but also and particularly those who did the counting. You probably imagine this latter group to be like Charles Dickens’s character Mr. Gradgrind, “a man of facts and calculations” who cared for little else. Popular culture has not, on the whole, embraced the 2009 prediction of Google’s chief economist Hal Varian, that statistician would be “the sexy job in the next ten years.” If statisticians have failed to dispel their unsexy image, then government statisticians have done even worse in the popular imagining. They are the grayest of the gray-suited bureaucrats, armed with notebooks, punch cards, calculators, or laptops, as the era allowed. (Arguably they should shoulder some of the blame: in a move entirely worthy of Dickens, the US Census Bureau has been located, since 1942, in a suburb outside Washington called Suitland.)29


  • "When we hear census, we think of numbers and statistics. But Andrew Whitby shows that the history of the census is an amazingly fascinating and illuminating story, and in The Sum of the People, he tells that story eloquently and persuasively. A real page-turner!"—Viktor Mayer-Schönberger, coauthor of Big Data
  • "In The Sum of the People, Andrew Whitby tells a gripping tale of humanity, civilization, and power. If you never imagined that a book about the census and the statisticians who conduct it could be a page-turner, think again. At a time when the need for the census is being challenged amid a tide of online big data, this book is also a deeply thought-provoking read."—Diane Coyle, author of GDP: A Brief but Affectionate History and Bennett Professor of Public Policy, University of Cambridge
  • "Humans spend much effort counting themselves. Always have, always will. Why? To control, conscript, and tax; but, then, also to hold accountable the powerful people who control, conscript, and tax. Andrew Whitby, alert to this duality, instructs and entertains as he brilliantly travels across the census landscape. Literally, a tour de force."—Kenneth Prewitt, Carnegie Professor of Public Affairs, Columbia University, and former director of the US Census Bureau
  • "This is a wonderful book. The history of the census may not at first appear to be a particularly hot topic, but Andrew Whitby's vigorous style, fine story-telling, and detailed knowledge combine to form a riveting narrative. Who would have thought that simply counting people could be such a deeply contested issue?"—David Spiegelhalter, author of The Art of Statistics

On Sale
Mar 31, 2020
Page Count
368 pages
Basic Books