The New Deal Claims the River

The conflict was not resolved. The conflict was the history. In Johnstown, Pennsylvania, in the summer of 1933, that history had a physical address: the second floor of the municipal building, where the city engineer’s office held a set of channel-depth surveys stacked beside a budget ledger. The ledger showed what the city could spend on the Little Conemaugh and Stony Creek in a single fiscal year. The surveys showed what the rivers needed. The difference between the two figures was the conflict, expressed in cubic yards of silt and dollars.

The channels through Johnstown had been dredged after the flood of 1889. They had been dredged again after 1894, and again after 1903. Each dredging removed what each subsequent freshet replaced. The 1933 surveys showed the channels constricted again. Sediment had built along the Stony Creek invert. Bridge piers that had been clear in 1910 now caught debris in ordinary spring runoff. Channel capacity was below what the 1903 improvements had provided. The city council had appropriated funds for maintenance. The funds covered roughly a quarter of the work the surveys specified. The engineer marked the shortfall in the margin of the ledger.

Two years later, in the spring of 1935, a different set of engineers arrived in the valley. They carried blueprints. They worked for the Works Progress Administration, the federal agency created that May. The blueprints showed channel widening, concrete retaining walls, debris basins, and levee construction along both rivers. The plans represented a scale of intervention the city had never attempted. They represented a scale of spending the city had never imagined. The federal engineers laid the plans on a table in the courthouse. The county commissioners looked at them. The city engineer looked at them. The plans were comprehensive. The plans were expensive. The plans were federal.

The contrast between the municipal ledger and the federal blueprints was the contrast between two conceptions of the same problem. The city engineer saw the rivers as a local burden. The city maintained the channels because no one else would. The city hoped the channels were adequate because the city could not rebuild them. The federal engineers saw the rivers as components of a regional system. The Little Conemaugh drained a watershed that crossed county lines. The Stony Creek drained a watershed that ran to the Conemaugh Gap. The gap was the constriction point where the combined flow of both rivers compressed into a channel too narrow for a major flood. The gap was the reason Johnstown flooded. The gap was not in Johnstown. The gap was in a different county. The federal engineers could cross the county line. The city engineer could not.

Forty-four years had passed since the South Fork Dam collapsed on Friday, May 31, 1889. The dam, located on the south fork of the Little Conemaugh River, fourteen miles upstream of the town, had been erected by the Commonwealth of Pennsylvania between 1838 and 1853 as part of the Main Line of Public Works, the state’s cross-state canal system. The state had sold it. The Pennsylvania Railroad had bought it. The railroad had sold it to speculators. The speculators had sold it to the South Fork Fishing and Hunting Club.

The club had cut down the embankment to widen a carriage road, lowering the dam by three feet. The club had allowed the spillway to clog. The club had packed breaches with mud and straw where stone and puddled clay had been. The dam had failed.

The water had killed more than two thousand people. The lawsuits had failed. The club’s members had paid nothing. The valley had absorbed the cost.

Now, in 1933, the city engineer stood at his desk and looked at a budget that could not make the rivers safe. The city had no authority over dams upstream. The city had no authority over the watershed. The city had jurisdiction over the channel banks within its borders. The city could dredge. The city could not build. The city could remove silt. The city could not prevent the silt from arriving. The instruments of municipal safety were a dredge, a budget, and a hope that the rain would not exceed what the channels could carry. The hope was not an engineering standard. The hope was all the city could afford.

Sixty miles to the southwest, in Pittsburgh, the offices of the federal engineering agencies were expanding. The Public Works Administration, created under the National Industrial Recovery Act of June 1933, had begun allocating funds to river projects across the country. The Army Corps of Engineers had surveyed the Ohio River basin, including the Allegheny, the Monongahela, and their tributaries. The Little Conemaugh was a tributary. The Stony Creek was a tributary. The Conemaugh River formed at their junction, in the center of Johnstown. The Corps had jurisdiction over navigable waters. The Little Conemaugh was not navigable. The Stony Creek was not navigable. Jurisdiction was the question.

The question moved through the federal bureaucracy in the winter of 1933 and the spring of 1934. The Federal Emergency Relief Administration had funds for employment projects. The Civil Works Administration had funds for construction. Each program had a different mandate. Each had a different ceiling. Each had a different definition of what counted as a worthwhile project. Flood control on a non-navigable tributary did not fit neatly into any of them. The Corps could study the watershed. The Corps could not build on a non-navigable stream without congressional authorization. The authorization required a river and harbor act. The act required a committee report. The committee required a survey. The survey required an appropriation. The appropriation required the authorization. The circle was the federal process.

The circle tightened in 1935. The WPA opened its Pennsylvania division. The division’s engineers carried blueprints into the Johnstown valley. The federal presence carried a logic that was not the valley’s own. The New Deal’s crisis logic had three components. The first was employment. Men needed work. The valley had unemployed men. The channel work would employ them. The second was infrastructure. The country needed public works. The valley needed flood control. The channel work would provide it. The third was national recovery. Federal spending in the valley would stimulate the local economy. The three components were national. The valley’s specific history—the memory of 1889, the record of repeated flooding, the knowledge that a dam upstream had killed two thousand people—was not a component of the logic. The valley’s history was the justification. The logic was the mechanism.

The mechanism absorbed the valley’s moral debt. The South Fork Club’s members had built a private retreat on a public dam. They had cut down the embankment. They had let the spillway clog. They had ignored warnings. The dam had failed. The courts had said the club was not liable. The valley had paid. Now the federal government was arriving with an appropriation and a blueprint. The appropriation was not compensation. The blueprint was not a verdict. The federal engineers were not settling a moral account. They were building infrastructure.

The distinction mattered. The legal system had asked whether the club owed the valley. The administrative system asked whether the valley needed the work. The two questions were not the same. The administrative system answered its question and moved on. The legal system’s question remained unanswered. The valley got concrete walls. The valley did not get justice. The concrete was the substitute for justice.

The ensemble of actors in the valley reflected the shift. The city engineer stood at his table with his surveys and his insufficient budget. The federal engineers stood at the courthouse with their blueprints and their appropriation. The county commissioners stood between them, approving the federal plans while continuing to appropriate municipal funds for local maintenance.

The residents of the valley stood in their yards and watched the survey parties stake the channel widening. Some residents welcomed the work. The work meant jobs. The work meant walls. The work meant the river would be wider and deeper and less likely to rise into their houses.

Other residents watched with a wariness that the memory of 1889 had taught. The federal engineers were from outside. The federal engineers would leave when the project was finished. The walls would remain. The maintenance of the walls would become the city’s responsibility. The city’s budget was insufficient. The insufficient budget was the reason the channels were constricted. The constricted channels were the reason the federal engineers had come.

The cycle had a structure. Private owners had neglected the dam. The dam had failed. The public had paid. The public had maintained the channels. The public’s maintenance had been insufficient. The channels had constricted. The federal government had arrived. The federal government had built. The federal government would leave. The public would maintain what the federal government had built. The public’s maintenance would be insufficient. The cycle was the deferred-maintenance debt, now nationalized. The debt had originated in a private ledger. The South Fork Club had kept a ledger of dues and improvements. The ledger had not included the cost of adequate repairs. The cost had been deferred. The deferral had passed to the valley. The valley had passed it to the city. The city had passed it to the federal government. The federal government had accepted it. The acceptance was the New Deal’s claim on the river.

The claim was specific. The WPA’s Pennsylvania division allocated funds for the Johnstown channel project in 1935. The project included concrete retaining walls along the Little Conemaugh through the city. It included debris basins upstream. It included levee construction along the Stony Creek. It included channel widening at the Conemaugh Gap. The project employed men from the valley. The project used federal money. The project was not a dam. The project did not address the dams upstream on the Little Conemaugh. The project addressed the channels. The channels were the receiving end of the watershed. The dams were the sending end. The federal government was working on the receiver. The sender was still in private hands or state hands. The sender was still subject to the same neglect that had destroyed the South Fork Dam.

The project was a partial answer. The partialness was the limit of the New Deal’s claim. The claim extended to the river. The claim did not extend to the dams on the river. The distinction between what the federal government could build and what it could regulate was the distinction between channels and dams. The federal government could build channels. Dam regulation was a state function. The state had regulated the South Fork Dam by not regulating it. The state was still not regulating dams effectively. The federal government was building channels. The channels would carry whatever the dams released. If the dams released too much, the channels would carry too much. The channels were wider. The channels were deeper. The channels were not infinite.

The Pennsylvania Department of Forests and Waters had jurisdiction over state dams. The department had inspectors. The inspectors visited dams. The inspectors wrote reports. The reports went to Harrisburg. Harrisburg sent copies to the dam owners. The dam owners responded. The responses went into files. The files accumulated. The accumulation was the state’s dam-safety program.

The program existed because of the South Fork Dam. The program’s capacity was limited. The program’s inspectors were few. The program’s authority to compel repairs was uncertain. The program could recommend. The program could not enforce.

Enforcement required legal action. Legal action required the attorney general. The attorney general had to decide whether to sue. The decision was political.

The politics of suing a dam owner were the politics of the public-trust principle that had emerged in the decades after the flood. The principle held that the public had an interest in the dam. The principle did not give the public the power to force the owner to fix it.

The gap between the interest and the power was the gap the state could not close. The gap was where the deferred maintenance lived.

The federal engineers in the valley did not concern themselves with the gap. The federal engineers had a project. The project had a budget. The budget had a timeline. The timeline had a completion date. The completion date was the date the federal government would leave.

The departure would transfer the maintenance of the new channels and walls to the city. The city’s budget was insufficient. The insufficiency was known. The federal engineers knew it. The city engineer knew it. The county commissioners knew it. The knowledge did not change the transfer.

The transfer was the structure of federal aid. The federal government built. The local government maintained. The local government could not maintain what the federal government built. The local government would try.

The trying would be insufficient. The insufficiency would accumulate. The accumulation would be the deferred-maintenance debt, passed from the federal ledger to the municipal ledger.

The municipal ledger was where the debt had always landed. The South Fork Club’s ledger had been private. The valley’s ledger had been public. The federal ledger was a temporary intervention in a permanent condition. The condition was that the rivers needed more maintenance than any single level of government could afford.

The residents of the valley lived inside the condition. A homeowner on Clinton Street, whose house stood within the floodplain the 1889 water had covered, watched the surveyors stake the line for a new concrete wall along the Little Conemaugh in the summer of 1936. The wall would be four feet higher than the existing bank. The wall would hold the river in a flood.

The wall was a federal project. The homeowner had paid taxes to the federal government. The taxes had become the wall. The wall was the return on the taxes. The return was partial.

The wall protected the house from a flood of a certain height. The wall did not protect the house from a flood that exceeded that height.

The height was an engineering standard. The engineering standard was based on the largest flood of record. The largest flood of record was 1889.

The standard said the wall would hold everything up to the 1889 flood. The standard did not say the wall would hold a flood larger than 1889. A flood larger than 1889 was possible.

The possibility was the residual risk. The residual risk belonged to the homeowner. The homeowner did not choose the standard. The federal engineers chose the standard. The choice was based on the record. The record was the history. The 1889 flood was the worst known. The worst known was the basis for the design.

The design was rational. The rationality did not eliminate the risk. The risk remained with the homeowner.

The homeowner’s taxes had paid for a wall that reduced but did not remove the risk. The reduction was the federal contribution. The residual was the homeowner’s burden.

The burden was the price of living below a dam. The price had been paid in 1889. The price was still being paid.

The payment had changed form. The form was now a concrete wall and a property tax.

The concrete wall was federal. The property tax was local. The local tax maintained the wall. The maintenance was insufficient.

The federal project was half-built in the autumn of 1936. The concrete walls along the Little Conemaugh were partially finished. The debris basins upstream were staked but not excavated. The levee along the Stony Creek was incomplete. The Conemaugh Gap widening was still on paper. The WPA had allocated funds. The funds were committed. The commitment was not the same as completion. The completion depended on continued appropriations. The appropriations depended on Congress. Congress depended on the administration’s priorities. The administration’s priorities could shift. The shift would leave the project unfinished. An unfinished flood-control project was not the same as no project. An unfinished project provided partial protection. Partial protection was less than full protection. The difference between partial and full was the margin in which a flood could occur. The margin was the residual risk. The residual risk belonged to the valley.

The concrete wall along the Little Conemaugh stood four feet above the old bank level. The wall was smooth. The wall was straight. The wall was federal.

The wall ended at a property line where the federal easement stopped. Beyond the property line, the bank was earth. The earth was the old bank. The old bank was lower.

The lower bank was the city’s responsibility. The city had no funds to raise it. The wall ended. The earth began.

The river did not notice the difference. The river would rise in the next flood. The river would find the low point. The low point was where the wall ended.

The wall ended because the federal appropriation ended. The appropriation ended because the project boundary ended. The boundary was on a map. The map was in a federal office. The office was in Washington.

Washington was far from the valley. The valley was below the dams. The dams were above the valley. The river connected them. The water would come down. The water always came down.

The valley depended on half-built federal works and the maintenance promises of distant agencies. The distant agencies had their own ledgers. The ledgers had their own entries. The entries did not include the dams.