The count of the dead began before the dead were fully located.
After a flood this large, recovery and counting proceed together, each incomplete without the other.
* * *
Palm Beach County bore the administrative weight of the south shore disaster. The county seat was West Palm Beach, on the coast, which had survived the storm with wind damage but not the catastrophic flooding that had struck the interior. The distance from West Palm Beach to Belle Glade was forty miles, across the muck plain that was still largely underwater on September 17th.
The routes that were passable in the first days circled the disaster zone rather than crossing through it. The count, in those first days, was assembled from the reports of whoever could reach the affected areas — the railroad workers who traveled the elevated embankments, the National Guard units dispatched in the first twenty-four hours, the volunteers who came with boats from the coast.
The official count, as it accumulated, distinguished between white and Black dead. That was consistent with the period’s disaster recordkeeping. What the racial columns revealed, as the count grew, was the distribution of death across the muck plain: Black entries concentrated in the labor camps and low quarters of the south-shore towns, white entries spread more broadly with higher survivor proportions relative to those present.
The count was not just a count of the dead. It was a map of who had been where.
* * *
Against this official count, each foreman on the south shore was conducting his own.
A foreman is expected to know, at all times, how many men are in his gang. He pays them, which means he holds their names. The daily count is a management function, not a humanitarian one, but its effect is that the foreman is the only person in the disaster zone who knows the specific roster of the men he brought from Georgia, from Alabama, from the Carolinas.
After the flood, the foremen began locating, accounting for, confirming the survival of each man. Some of the men were confirmed dead and their bodies recovered. Some were confirmed dead and their bodies were not recovered — carried into the Everglades by the water that moved south and east, into the sawgrass and the sloughs, into the country that did not give back what it received.
Some of the men were neither confirmed dead nor confirmed alive.
The foreman’s count, in these cases, did not close. Not dead by confirmation. Not alive by confirmation. Simply absent from the ledger.
* * *
The death toll of the Okeechobee flood is not a fixed number.
The official count in the immediate aftermath: approximately 1,800. The revised count developed by subsequent historical research, incorporating all available sources: approximately 2,500 to 3,000.
The space between these numbers is the space that the counting systems of 1928 could not close.
The methodological problem with the official count was not carelessness. It was that the population being counted had characteristics that made counting difficult. The migrant agricultural labor force that populated the south shore during harvest season was not registered. There was no roster of names, no record of who had arrived, from where, under whose employ. A man who had come from Georgia to pick beans in the Okeechobee muck existed, as a countable entity, only in the foreman’s ledger — and if the ledger was destroyed in the flood, or the foreman’s account was incomplete, the man did not exist in any recoverable system.
If his body was not found — if the water had carried it into the Everglades — he did not become a confirmed death. He became an absence in the count.
The portion of the death toll that remains most uncertain — the upper range of the revision — is also the portion that was most heavily Black. The county coroner’s records, the death certificate system, the newspaper obituaries, the Red Cross relief lists: each captured different subsets of the dead, and each captured the Black agricultural laborer population less completely than the white property-owning population.
* * *
The body recovery from the Everglades presented the count with its most persistent uncertainty.
The water that moved south and east from the lake carried the flood into the northern Everglades — into the sawgrass and the wet prairies and the shallow sloughs of the country south of the muck plain. Recovery teams entered the Everglades in the weeks after the flood and worked in conditions that were difficult to overstate: no roads, subtropical heat, a landscape that looked the same in every direction.
The bodies they found had been there for weeks in that heat.
The bodies they did not find remained in the Everglades. The vegetation, the water, the soil work on organic matter at their own pace. A body in the northern Everglades in October 1928 could be in a location that a recovery team twenty feet away would not find.
The foremen who had kept their gangs on the south shore had, after the flood, the same problem at a human scale. The men unaccounted for were somewhere. Whether they were in the muck, in the Everglades, or alive somewhere and simply gone — this was a question the available information could not answer.
A laborer from Georgia whose name was in a foreman’s ledger and was not crossed off after the flood existed, for his family, in a state of unresolved absence. He might have survived and not returned. He might be dead in a location no one had reached. The absence held all three possibilities simultaneously until it held none of them, which it might never become.
* * *
The civilization the Okeechobee entry measures — the civilization of agricultural infrastructure and canal systems and railroad embankments and county governments — had no mechanism designed to resolve this category of absence for the families it concerned.
There was no system for notifying the family in Georgia that a man had not returned from the south shore. The systems that existed were local. The disaster was local. The people who had died were not.
They had come from elsewhere, and their absence was felt elsewhere, in places where the count could not be sent.
The Archivist held the count at the range the available evidence supports. He held the space between the numbers as the space the counting systems of 1928 could not close. He held the foremen’s open ledgers beside the official count as entries of equal weight.
The people who were not counted were not absent from the flood.
They were absent from the count.