In the summer of 1927, Ora Marshino of the United States’ Children’s Bureau wrote to officials in all participating states asking them to better account for their use of federal funds distributed through the Sheppard-Towner Act. The states’ directors, it seemed, had a great deal of trouble using the forms and reporting procedures that the Children’s Bureau had devised. As Marshino wrote to the Chief of Pennsylvania’s Preschool Division under the State Department of Health, “whoever figured up your budget for you, seems to have a different system of mathematics from ours.”Footnote ¹ Embarrassed, the state-level bureaucrat sheepishly agreed that “our arithmetic was unique.”Footnote ² In response to the same inquiry, Kentucky’s bureau director replied that the accounting Marshino asked for did not reflect her state-level work and claimed it was like “a Chinese puzzle” to try to fit her state’s statistics into the form the federal bureau requested.Footnote ³ New York’s bureau director likewise replied, “I think if I see another [federal] questionnaire this year I shall curl up and die.”Footnote ⁴

These state officials were neither truculent nor incompetent, but they were befuddled by the new reporting imperatives created by federal matching grant programs. Unaccustomed to correlating state-level statistics with other states for use by a federal agency, these bureaucrats’ inability to comply with requests for federal data captures the shifting relationship between federal and state agencies in the early twentieth century. As such correspondence demonstrates, the fact that the Children’s Bureau staff had to rely on state administrators for data about federally-funded programs hindered their efforts to collect standardized national statistics. Ultimately, this barrier contributed to Sheppard-Towner’s defeat. As this article will show, the Bureau staff learned from administering Sheppard-Towner that they needed to implement more stringent federal oversight over state-level accounting in subsequent maternal and child health laws.

Colloquially known as the “Better Baby Bill,” the Sheppard-Towner Act of 1921 was Congress’s first attempt at creating a universal social welfare and healthcare program.Footnote ⁵ It was also an early federal matching grant. The law provided federal funds to the states for infant and maternal welfare programs—well-baby clinics, midwife training and regulation, birth-registration campaigns, and prenatal and infant-wellness education—until 1929, when Congress declined to renew its funding. Though the law was short-lived, the problems the Children’s Bureau encountered administering it provide insights into how federal matching grant programs began to shape federal and state relations before the New Deal.

Though it was one of a number of federal matching grants to the states in the 1920s, the Sheppard-Towner Act is uniquely instructive.Footnote ⁶ As political scientist Kimberley Johnson has argued, unlike the USDA or the Bureau of Chemistry (that also administered federal matching grants at the time), the Children’s Bureau lacked the broad coalitional and institutional support that other federal bureaus enjoyed. This made the Children’s Bureau staff and its flagship legislation distinctly vulnerable to accusations of federal overreach.Footnote ⁷

Rightly weary of such charges, Bureau staff trod lightly around issues of state authority over programming. Though Bureau leadership wanted to model Sheppard-Towner after nationalized maternal and infant health policies abroad, the politics of early twentieth-century federalism forced them to grant state directors wide administrative latitude in programming and accounting. This emphasis on state autonomy in administration impeded one of the Bureau’s primary objectives—to collect reliable national and state-level statistics about infant and maternal mortality and to identify which programs ameliorated them.

Allowing states ultimate authority over programming and accounting meant that each state director documented their Sheppard-Towner work differently—if they did so at all. When it came time to renew federal appropriations in Congress, the Bureau staff found themselves unable to create and present national statistics on the workings of state programs. Without reliable national data, the Bureau’s agents had little ability to defend their use of federal dollars when the political winds shifted in Washington and the law came under more organized attack. The Bureau’s staff’s inability to prove the worth of Sheppard-Towner programs with numbers gave skeptical lawmakers a convenient excuse to defund the law in 1929. Unwilling to be caught in the same political trap twice, the Bureau’s leadership allocated themselves more administrative control over state-level data collection in subsequent legislation.

Focusing on the role of information-gathering in the administration and defeat of the Sheppard-Towner Act sheds new light on the political history of this watershed legislation. Historians of women and gender have long taken interest in the U.S. Children’s Bureau—the first female-run federal agency in the United States—and its legislative agenda. Motivated to uncover the history of women interacting with and shaping the federal government, such scholars have uncovered the rich political history of Progressive Era women’s associational networks that wielded significant political pressure in Washington and in state houses in the early twentieth century.Footnote ⁸ The creation of the Children’s Bureau and the passage of Sheppard-Towner are both signal accomplishments in this history.

At the same time, historians have also concentrated on the female-run Children’s Bureau’s gendered political conflict with the American Medical Association (AMA), the Public Health Service (PHS), and antisuffrage organizations. In so doing they have used Sheppard-Towner to explain not just the rise but also the fall of women’s political influence at the federal level in the 1910s and 1920s.Footnote ⁹ However, primarily gendered analyses of the legacy of the law de-emphasize the structural and institutional factors that also hastened its demise. When we examine the legacy of the Sheppard-Towner Act through the analytical lens of American political development and state-building, the law’s fate exposes new contours in the evolution of federal and state administrative capacity.Footnote ¹⁰

More recently, economists and political scientists have centered Sheppard-Towner in their discussions of the rise of early twentieth-century “New Federalism” and maternal and infant welfare spending at the state level.Footnote ¹¹ Such scholarship has done a great deal to uncover the political and structural impediments to Sheppard-Towner’s longevity. This article adds to this body of work by pointing to federal data collection as a central hurdle that the Children’s Bureau faced as it tried to administer an early matching grant program through the states. Additionally, this piece shows how institutional capacity was shaped over time by linking Sheppard-Towner’s political failure to the subsequent administration of the Social Security Act.Footnote ¹²

When we shift focus away from gender and toward the politics of American federalism and data collection, the law proves a transitional piece of legislation that can help explain how the nineteenth-century “government out of sight” became the vast expansion of federal social welfare programs under the New Deal.Footnote ¹³ Though the history of Sheppard-Towner is usually told as a story of defeat, the Children’s Bureau staff retained the administrative lessons they learned through their battle over the law and incorporated them into their programs under the Social Security Act. Even in its political failure, Sheppard-Towner was (and is) instructive.

Scholars have focused on the gendered history of the Sheppard-Towner Act for good reason. The bill was the product of a watershed moment for American women in politics. It was the brainchild of the Chief of the Children’s Bureau, Julia Lathrop, the first woman to run a federal agency. The nation’s first congresswoman, Jeannette Rankin (R-MT), proposed Lathrop’s bill to Congress in 1918. Congress passed the final version of the law just one year after the ratification of the Nineteenth Amendment. There is more to the story of the law than only as an early legislative victory for women, however. It can also teach us about the rise of data collection at the federal and state level and how the politics of American federalism subverted bureaucrat’s efforts to understand their constituents through numbers.

The Children’s Bureau was founded in 1912 with the express mandate to collect and distribute information about the nation’s children. These aims were clearly outlined in the Act establishing the Bureau. The law stated “that said Bureau shall investigate and report upon all matters pertaining to the welfare of children and child life among all classes of our people.”Footnote ¹⁴ This emphasis on investigating, synthesizing, and reporting suited the staff of the Children’s Bureau well. Most of the original staff had worked in settlement houses, many in Chicago’s Hull House. While there, they were trained as or by social scientists to value social and vital statistics and to believe in the power of data to ground policy that would improve the lives of immigrants and the working class.Footnote ¹⁵

The early work of the Bureau reflects this belief in the power of data to transform an unequal society. From the Bureau’s founding until the 1920s, the agency’s primary work was gathering information and creating national standards by which they could measure their work. Just as their founding mandate required, Bureau staff in the 1910s undertook birth registration campaigns, investigated the causes of infant mortality in communities, and created national standards for weighing and measuring infants. They publicized the information they collected through the popular press and with publicity drives like national Baby Weeks in 1916 and 1917 and the Children’s Year in 1918.

As Bureau staff began to view their mandate more broadly, they advocated for a legislative agenda to address some of their findings. The staff chose to focus on reducing national rates of infant mortality, thinking this the least controversial agenda within their domain. To promote legislative action, the staff used newly available vital statistics from the military to dramatize their findings, arguing that the army would have a more robust infantry if only the nation invested in the health of its small children.Footnote ¹⁶ Likewise, they contrasted America’s high rates of infant mortality to other nations, hoping comparative graphs of vital statistics would make the case for federal intervention. Arguing that national infant mortality rates were a valuable index of national health, Bureau staff promoted expanding the birth registration area and promoted public health campaigns aimed at mothers and young children. In so doing, Bureau staff hung their legislative aspirations on reducing the infant mortality rate.

Publicizing and dramatizing the national infant mortality rate gave Bureau staff and their allies a concrete—and to their minds, indisputable—way to promote their legislative agenda. However, the timing of the bill’s final passage was also significant. Historians agree that advocates of Sheppard-Towner found the political support they required by riding the coattails of the Nineteenth Amendment. Congress voted on a version of the maternity and infancy bill three times before President Warren Harding signed it into law in November 1921. Because women were politically mobilized around suffrage and because politicians feared punishment at the ballot box if they voted against a bill so popular with their new constituents, the bill found overwhelming support in Congress in 1921.Footnote ¹⁷ Though the law sailed through Congress, many legislators remained privately skeptical. One senator reportedly confessed that “if the members could have voted on the measure secretly in their cloak rooms it would have been killed as emphatically as it passed.”Footnote ¹⁸

Politically precarious from the start, the final version of Sheppard-Towner only provided funding for five years, requiring Congress to reevaluate the programs in 1927.Footnote ¹⁹ Additionally, the final version of the law contained contradictions about the Children’s Bureau’s role in administration. The law ostensibly gave Bureau staff the ability to collect data from state directors, stating that “it shall be the duty of the Children’s Bureau to make or cause to be made such studies, investigations, and reports as will promote the efficient administration of this Act.”Footnote ²⁰ However, it also forcefully stated that “this Act shall be construed as intending to secure to the various States control of the administration of this Act within their respective States, subject only to the provisions and purposes of this Act.”Footnote ²¹ The tension between the Bureau’s need for efficient national data collection and the political need for state autonomy was written into the legislation itself.

These contradictions reflected not only the political precarity of the law but also mirrored Julia Lathrop’s competing ideas for how best to carry out a program to promote maternal and child health. Lathrop modeled much of her vision for the law after national programs underway in Britain and New Zealand.Footnote ²² However, local volunteerism had been the backbone of the early Children’s Bureau initiatives that Lathrop ran—the bureau’s first birth-registration campaigns, infant and maternal mortality surveys, and Children’s Weeks. As Theda Skocpol has shown, Lathrop’s plan for the bill thus “straddled the boundary between a comprehensive national health program, on the one side, and localism and volunteerism on the other.”Footnote ²³ Following the precedent of earlier matching grants such as the Smith-Lever Act, the Sheppard-Towner Act required states to opt in by accepting the provisions of the law in their legislatures and appropriating state funds to match the allotted federal dollars. The law and its chief administrator reached for states’ autonomy and local control with one hand and for national standards and centralized bookkeeping with the other.

The law did stipulate that the Children’s Bureau and its advisory Maternity Board (made up of the Chief of the Children’s Bureau, the Surgeon General, and the Commissioner of Education) had the authority to withhold funds if they believed a state had used previous years’ monies irresponsibly. However, the political stakes of appearing to threaten state sovereignty were too high. Lathrop’s successor as Chief of the Bureau, Grace Abbott, trod lightly around issues of state sovereignty and rarely even checked the states’ financial reports against their proposed annual budgets. Not once in the law’s tenure did the Bureau threaten to withhold annual funding, though Bureau employees believed that state directors routinely misused the funds.Footnote ²⁴ Without the political capital to withhold funding for state programming, Bureau staff had no mechanism to ensure that states made proper use of federal dollars or accounted for their work.

The Bureau leadership was right to worry about political attacks of all kinds. Though lawmakers wrote restrictions to federal authority into the law, Sheppard-Towner continued to unnerve many politicians, professional groups, and private citizens. Religious institutions, branches of the American Medical Association, and antisuffragists contested the constitutionality of the Sheppard-Towner Act repeatedly over the course of the 1920s. Accused of overriding individual parental and religious rights, fomenting a Bolshevik plot to turn the United States socialist, and promoting birth-control methods and free love, the Children’s Bureau weathered a near-constant storm of criticism while administering Sheppard-Towner funding. However, the rhetoric of federal overreach proved the most potent political critique available to detractors of all stripes.

Shortly after the bill’s passage, the state of Massachusetts and several individual Massachusetts residents sued the Secretary of the Treasury, the Chief of the Children’s Bureau, the Surgeon General, and the Commissioner of Education over the constitutionality of the Sheppard-Towner Act. The plaintiffs appealed the case to the Supreme Court in May 1923, arguing that the Act violated the Tenth Amendment. Representatives of Massachusetts argued that the law forced wealthy industrial states to pay an undue share through the federal income tax. The plaintiffs in Massachusetts v. Mellon also took their logic one step further. They not only argued that the Sheppard-Towner Act was unconstitutional but claimed that all federal matching grants were.Footnote ²⁵

The Supreme Court ultimately dismissed the cases, declaring they “must be disposed of for want of jurisdiction without considering the merits of the constitutional questions.”Footnote ²⁶ However, the justices went on to write in their ruling that Massachusetts’ argument did not hold because “nothing has been done and nothing is to be done without their consent.”Footnote ²⁷ Federal matching grants therefore stood as constitutional because states were not forced to accept the provisions of the law and could opt out—as Massachusetts ultimately did. The issue, the court declared, was political not judicial.Footnote ²⁸ The matching grant model that the Children’s Bureau adopted proved not only politically expedient but legally sound.

Even as the Children’s Bureau staff defended the constitutionality of the Sheppard-Towner Act by emphasizing state authority, both they and state directors showed a great deal of ambivalence about total state sovereignty over programming. Some of this ambivalence was, as we have seen, a result of the law’s contradictory mandates. In practice, the Bureau agents urged state directors to implement their specific and uniform model of programming. However, when state directors chose to defy the Bureau’s guidance—and most did to some degree—the federal agents did nothing.

State directors were often daunted by the task of running programs with both federal and state funds. They relied on the Children’s Bureau staff to help them sort out the possibilities and flooded the Bureau with logistical questions about how federal dollars could be spent. For example, Texas’s director wondered if he could use Sheppard-Towner funds to pay a cartoonist, presumably to illustrate promotional health materials.Footnote ²⁹ A Bureau agent said that it depended on the subject of the cartoons.Footnote ³⁰ Delaware’s bureau director asked if he could purchase automobiles for the state nurses to use while they traveled to far-flung communities, “and if so, will such machines belong to the State of Delaware or to the United States?”Footnote ³¹ A Children’s Bureau agent told him that his state likely could use the funds to purchase an automobile and she believed—though she was not certain—that the car would then belong to the state of Delaware.Footnote ³² Mixing federal and state funds created confusion over authority and ownership for bureaucrats at all levels of government.

While most state directors seemed to desire and seek out federal oversight and guidance in programming, some state directors understood that ultimately power over programming resided with them and freely ignored the Bureau agents’ wishes. For example, when Children’s Bureau agent Dr. Florence Kraker went to visit the state director in Arkansas, she discovered that his bureau was examining and treating school-age children in addition to infants. After Kraker admonished the state director, the director “acknowledged that he was acting contrary to the law” but defended his actions.Footnote ³³ He was simply doing what he felt was appropriate for his state, he argued.

Furthermore, he claimed he was not alone in defying the letter of the law. “We have made a rather careful investigation,” he explained, “and find that other States are carrying out the same policy.” “The probable difference,” between himself and other state directors, was that “we have been perfectly frank in this State.” “In the future,” he promised, “we shall not embarrass you by simply not reporting the school work.”Footnote ³⁴ Arkansas’ director’s open insubordination is revealing. He and Kraker both understood that the federal Bureau would take no action beyond again asking him to stop.

Most state directors ignored the advice of Children’s Bureau agents in more subtle ways. Nearly every state used Sheppard-Towner funds to support Better Baby Contests—competitions at state fairs where physicians and nurses scored babies on health and attractiveness.Footnote ³⁵ These were an exceptionally popular use of Sheppard-Towner funding at the state level because physicians and public health nurses could reach a wide swath of the population without having to travel far to do so. But the Children’s Bureau staff did not approve of the baby contests because they worried it encouraged families to bring only their healthiest babies to be examined by physicians, and therefore undermined Sheppard-Towner’s aim to reach the most vulnerable mothers and children.Footnote ³⁶ Bureau staff objections did little to discourage the contests at the state and county level, however.

In practice, both the Children’s Bureau staff and the state directors showed a great deal of ambivalence about total state control over programming. While some state directors bristled at even minimal federal oversight, others relied heavily on the Children’s Bureau to set up their programs and even staff their bureaus. Interactions between federal agents and state administrators reveal that while Bureau staff wanted to administer a uniform, national maternal and infant health program through the states, the political compromises they had made to pass the law forced them to defer to the states whenever conflict arose. As the Bureau staff would soon learn, their caution around state sovereignty made the task of measuring and accounting for state programs nearly impossible at the federal level.

Even as the Bureau agents tried to administer the law with deference toward state autonomy, they found themselves defending centralized federal administration in Congress. At the law’s reappropriations hearings in January 1926, congressional representatives made federal overreach the central issue. Grace Abbott, representing the Children’s Bureau and Sheppard-Towner, tried to carefully thread the needle between insisting that federal funding was necessary for these programs while also advocating federal restraint. When a member of Congress directly asked her, “Are you in favor of the Federal Government getting out at all?” she carefully responded, “I am in favor of the Federal Government getting out as soon as the situation is such that we will get the work properly done.”Footnote ³⁷ Unfortunately for Abbott, this line of reasoning proved unconvincing when the reappropriation hearings moved to the Senate.

Following the contentious congressional hearings that year, the Bureau hired Dr. S. Josephine Baker, founder of New York’s maternal and infant health bureau and prominent advocate for maternal and child health programs nationally, to compile a report on the success of Sheppard-Towner programs around the country. The initial five-year period of federal funding was set to end in 1927 and the Bureau leadership hoped that Baker’s report would demonstrate the success of Sheppard-Towner programs to reticent lawmakers. With hard statistical proof, they wanted to demonstrate the importance of federal intervention.

Unfortunately for the Children’s Bureau, the law’s limitations on federal authority made assessing programs at the national level nearly impossible. Baker began her study by sending out a survey to participating states asking how many infants and mothers Sheppard-Towner workers reached, how many state employees were paid with the funds, how much money was expended on each initiative, and how many lives the programs had saved. “What I am trying to get at,” Baker explained, “is some sort of a relationship between the amount expended in each state during the years that the S-T money has been available and the maternal and infant death rates during that time.” “In other words,” she wrote, “can we show a direct relation between the money so far expended and any reduction either in the maternal or infant death rate.”Footnote ³⁸ She quickly lost hope in her ability to do so, explaining that the further she looked into the matter, “the more I am impressed with its complexity. What seemed like a very simple matter at first now seems filled with all sorts of difficulty.”Footnote ³⁹ After corresponding with every participating state director, she was forced to admit that she had become “rather doubtful of obtaining any worth while [sic] actual information from the Directors at least any that we can use as a basis for a standard for all states.”Footnote ⁴⁰

Much to Baker’s chagrin, most state directors could not answer the questions on her survey either because they had not surveyed their own work or because they had collected information in a manner different from what the Children’s Bureau wanted. New York’s state director reported meticulously on the workings of the state’s maternal and infant health programs. Still she lamented, “It is certainly disappointing, after all the effort we have put in to get proper records established and in general use, to find that we cannot seem to produce the right kind of information, but it would seem that such is the case.”Footnote ⁴¹ Many state directors appended their own statistical reports to Baker’s questionnaire to show that they had taken their own measurements, but this information proved useless to Baker because each state director measured different things. Furthermore, Utah, New Mexico, and Arizona’s directors reported that they had no statistics on their Sheppard-Towner work at all.

Many state directors, nearly all of whom were doctors or nurses themselves, felt that their real work was treating and instructing their constituents, not accounting. New Hampshire’s director complained that it was too hard to do clerical work after traveling so much to reach the mothers and children of her state.Footnote ⁴² Arizona’s director likely spoke for many when she replied to Baker’s inquiry, “Good records are important it is true but I feel that we cannot consume too much valuable time analyzing the work done at the expense of the work itself.”Footnote ⁴³ In the end, Baker could only include results from fourteen states in her report. This was the irony of American federalism. Allowing the states authority over their own programs meant the Children’s Bureau agents had no way to measure Sheppard-Towner’s effectiveness and no way to defend their work in Washington.

The Children’s Bureau’s inability to create and collate their own vital statistics proved central to the growing political opposition against renewing Sheppard-Towner funding. Having used infant mortality as an index to advocate both for the founding of the agency itself and its flagship legislation, the Children’s Bureau leadership had implicitly promised data related to infant deaths in the United States. However, the actual data collection depended on the states to opt in to the birth registration area and the numbers themselves were tallied by the Bureau of Vital Statistics. Likewise, as Baker’s trouble compiling the 1926 report shows, the Bureau felt ownership over but did not have total access to state-level statistics because Sheppard-Towner was written to protect state sovereignty. This left the Bureau’s leadership in a precarious political position because they did not own the data it needed to justify their own programs.

The Children’s Bureau’s inability to account for the federal dollars used by the states was not lost on the law’s swelling ranks of detractors. While many historians have pointed to the American Medical Association’s growing opposition to Sheppard-Towner as the central crux of the law’s defunding, nearly all these accounts stress the organization’s opposition to state medicine.Footnote ⁴⁴ None have explained how the AMA packaged its opposition in the language of both federalism and data collection.

During Senate hearings in 1927, Senator Phipps (R–CO) read aloud a public letter from the AMA opposing the law. In the letter, the national association of physicians pointed repeatedly to the Children’s Bureau’s failure to present reliable data proving the benefits of Sheppard-Towner programs. “In recent hearings before the House Committee on Interstate and Foreign Commerce, as in all other propaganda in support of the Sheppard-Towner Act,” Phipps read, “one looks in vain for facts and figures showing a reduction in maternal and infant mortality through the operation of the act.” The AMA argued that the Children’s Bureau had presented Congress with statistics “of the most general kind, not properly correlated to Sheppard-Towner activities, too often from interested sources and not infrequently from persons who are hardly to be regarded as competent to speak on such matters.”Footnote ⁴⁵ The AMA even suggested that infant mortality had increased since the passage of the law in some states. Without reliable statistics to prove otherwise, the Children’s Bureau leadership could do little to refute such accusations.

While the AMA and their political allies in the Senate charged the Children’s Bureau with failing to properly oversee national uniformity in data collection, they simultaneously accused the Bureau staff of undermining states’ rights. They correctly noted that state supremacy was “conceded by the Sheppard-Towner Act itself, for through it the Federal Government seeks, not to force its way into the State health program but to pay the State for the privilege of supervising and directing it.” “[I]f the Federal Government can buy from the states the right to supervision and control of State health activities,” they argued “there is no reason why the Federal Government should not likewise buy the other constitutional rights of the States.” “It is to that end” they concluded, “that the Sheppard-Towner Act seems to lead. The accomplishment of that end will be coincident with the destruction of our present system of government.”Footnote ⁴⁶ The AMA’s two-pronged critique of Sheppard-Towner exactly exposed the contradictions of the law: the Children’s Bureau could not effectively run a uniform, national health program and allow states full sovereignty.

That same year, Children’s Bureau staff were forced to articulate their position on federal versus state authority for another unexpected reason. Following some bureaucratic and operative disputes with the Bureau, the U.S. Comptroller General authorized the national accounting office to withhold Sheppard-Towner funding to states indebted to the federal government for reasons unrelated to Sheppard-Towner. Though the Children’s Bureau staff and the federal board had already approved their annual proposals, the General Accounting Office refused to send Louisiana, Tennessee, and Montana’s Sheppard-Towner funding for the 1927 fiscal year.

Believing the Comptroller General’s actions were unwarranted and unprecedented, Children’s Bureau agents contacted the heads of the Public Health Service, the Bureau of Education, and the Federal Board of the Department of Vocational Education. They asked these other federal agents if the Comptroller General had ever withheld funds from their department’s matching grant programs. All three department heads replied that no payments had ever been withheld from the states to offset indebtedness or for any other purpose.Footnote ⁴⁷

Children’s Bureau agents then scrambled to come up with a response to the comptroller and to find money for the affected states. Ultimately, they landed on a line of reasoning that directly contradicted their defense of state sovereignty. They argued that federal dollars remained federal dollars, even when loaned to the states through matching grants.Footnote ⁴⁸ They took this logic directly from the Comptroller General himself who had previously argued that “the United States is the owner of [loaned] funds up to the point of actual expenditure … even while actually in the possession of the State.”Footnote ⁴⁹ Ultimately then, the Comptroller General’s fiscal roadblock paired with critiques from politicians and interest groups forced Children’s Bureau leadership to conclude that matching grant programs required federal agencies to take a more direct interest in state programming. If federal dollars remained federal dollars when loaned to the states, then the federal government retained ultimate authority over state programs.

From this incident onward, the Bureau took a much more active role in auditing state bureaus.Footnote ⁵⁰ While they could not force state directors to comply, Bureau staff began to send blank standardized reporting documents to each state director with the hope of getting state-level statistics they could collate.Footnote ⁵¹ The state directors complained but did their best to answer, though many still could not. By then, however, it was too late.

By 1927, enough election cycles had passed since the passage of the Nineteenth Amendment to reveal that women did not vote as a single block. This allowed politicians—many of whom had been skeptical of federal intervention all along—to publicly oppose renewing Sheppard-Towner. President Calvin Coolidge was among them. During reappropriation hearings in Congress, the president had a statement read into the record. “Federal interference in State functions can never be justified,” the president contended, because “[a]s shown in the maternity and infancy act, when the Government engages in such an enterprise it is almost impossible to terminate its connection therewith.” Coolidge implored the legislators to “not only decidedly refuse to countenance additional Federal participation in State-aid projects” but to “make careful study of all our activities of that character with a view to curtailing them.”Footnote ⁵² Using Sheppard-Towner as the prime example of federal excess and inefficiency, lawmakers in 1927 placed all federal aid to the states on trial alongside it.

Even after the president opposed them, the leaders of the Children’s Bureau did not lose hope in the viability of the matching grant model. They proposed a similar matching grant program, the Newton Bill, in 1928, just before Sheppard-Towner was set to expire. When revising drafts of the bill, Katherine Lenroot, then the Acting Chief of the Bureau, interviewed directors of other bureaus under the Department of Agriculture, who administered similar matching grant programs, but who faced little political pushback. Part of the reason for this, Lenroot discovered, was that the Department of Agriculture had a much more diffuse and interconnected web of federal support within the states. At the same time, under some bureaus’ programs, she learned, no money was ever given directly to the states: “the federal government simply pays its share of the expenses, whatever that may be” and therefore maintained more administrative control.Footnote ⁵³ After interviewing these bureau directors and having had the trouble with state agencies that she’d had with Sheppard-Towner, Lenroot concluded that her new bill would need to allocate her Bureau more administrative control over programming and accounting.Footnote ⁵⁴

Lenroot and her colleagues wrote these lessons into the new legislation. The proposed Newton Bill explicitly outlined the more active role the Children’s Bureau would take in its administration. State-level maternal and infant health work “shall be carried on in such a manner as may be mutually agreed upon by the Children’s Bureau and the State agencies receiving the benefits of this Act,” the final draft read.Footnote ⁵⁵ Furthermore, the bill hinted at a new federal oversight requirement by including stipulations about reporting.Footnote ⁵⁶ Though such revisions likely would have improved the Bureau’s ability to administer and defend the law, they did nothing to quell the criticism of federal overreach. In the end, the Newton Bill confronted the same political opposition that had thwarted Sheppard-Towner’s renewal.

As Children’s Bureau leadership continued to propose ill-fated bills to renew Sheppard-Towner work, they stressed the importance of designating themselves more administrative control than they had had under Sheppard-Towner. After viewing a draft of another proposed maternal and infant welfare bill in 1930, Katherine Lenroot explained to Grace Abbott that she could not support the bill in its current iteration “for the reason that there is not a clear division of administrative responsibility.”Footnote ⁵⁷ Administrative control over programming was now nonnegotiable from her perspective. Her agency needed access to iron-clad state-level statistics and accounting in order to defend against political attack.

Lenroot’s instincts proved correct when Children’s Bureau agents learned at the 1930 White House Conference on Children that President Hoover wanted to relocate the Children’s Bureau to the Public Health Service (PHS). The move seemed logical to the leaders of the PHS and President Hoover because the bulk of the Bureau’s purview in the 1920s had been public health work supported by Sheppard-Towner. The Children’s Bureau staff, however, hated this idea because they feared it would mean a loss of female control over infant and maternal welfare work and worried that without an independent bureau, such work would be further degraded and neglected.Footnote ⁵⁸

Significantly, many of the arguments made for and against moving the Children’s Bureau to the PHS centered on statistics and data collection. Just as opposition in the House and Senate had, PHS agents argued that the Children’s Bureau staff had not collected reliable statistics in their administration of Sheppard-Towner programs. “If you will examine the statistics presented by the Children’s Bureau,” a PHS representative argued, “you will discover how thoroughly futile has been their activities.”Footnote ⁵⁹ Additionally, the agent contended, the faulty statistics that the Bureau had collected “were already being carried on in the Public Health Service … and there was no need to duplicate that organization in another department.”Footnote ⁶⁰

Representatives of the Children’s Bureau responded by arguing that, unlike the PHS, their focus was the welfare of the whole child and that reducing infant mortality and promoting general health were important aspects of this agenda. Moreover, they claimed that the PHS studies had followed theirs, not the other way around: “If there has been any duplication it has been by the Public Health Service and not by the Children’s Bureau.”Footnote ⁶¹ However, they were also left to defend their sparse data on Sheppard-Towner programs. “Rates alone over such a short period of the time cannot be used as a measure of the value of the work done,” a Bureau representative explained.Footnote ⁶² Neither agency acknowledged that the Children’s Bureau was not solely responsible for collecting vital statistics on the infant mortality rate or that states were responsible for their own program’s accounting. The Children’s Bureau had staked its claim on these numbers and the numbers did not satisfy.

In the end, the Children’s Bureau remained under the Department of Labor and cooperated with the PHS in the administration of subsequent maternal and infant health policy. As other historians have noted, the conflict at the 1930 White House Conference on Children between the Children’s Bureau and the PHS had tensions over gender and authority simmering just below the surface.Footnote ⁶³ However, the arguments presented by both sides also reveal a turf war over vital statistics and state programming.

When Sheppard-Towner funding lapsed in 1929, responsibility for funding the maternal and infant health programs fell back to the states. Without the infusion of federal dollars and programmatic assistance from the Children’s Bureau, most state programs deteriorated. As one physician wrote to the chairman of the House Ways and Means Committee in 1935, “When the federal support, offered by the Shepherd-Towner Act, was withdrawn, almost all states … stopped or reduced markedly the maternal and child welfare activities. This is regrettable.”Footnote ⁶⁴ Contemporary economists have confirmed this physician’s assessment. Carolyn Moehling and Melissa A. Thomasson have shown that without the federal matching grants, state investment reverted back to pre–Sheppard-Towner levels. States that had invested heavily in maternal and infant health programs before federal funds were made available continued to do so; states that had not stopped funding the programs altogether.Footnote ⁶⁵

In the years between the lapse of Sheppard-Towner in 1929 and the 1935 passage of the Social Security Act, Children’s Bureau staff busied themselves with collecting state-level and national statistics. Unwilling to be caught empty-handed again, Bureau employees spent the intervening years collecting data. They researched and compiled reports on the economic crisis, lapsing funding for mother’s pensions at the state level, and the impact that state-level budget cuts had on public health initiatives for children.Footnote ⁶⁶ When the political moment was right, Bureau staff would have the data they needed to propose new legislation.

The Children’s Bureau leadership tried annually to revive federal support for infant and maternal health and welfare programs. They found little political support until the New Deal coalition took power. With the support of the Roosevelt administration, Bureau staff were able to once again fund Sheppard-Towner programs under Title V of the Social Security Act. As Kimberley Johnson has noted, the Bureau leadership’s strategy to leverage their statistical resources gained them "a seat at the policy table" within the new administration.Footnote ⁶⁷

Children’s Bureau agents wrote much stricter federal regulations about state-level accounting into Title V than they had into Sheppard-Towner. The Bureau staff administered three parts of the Social Security Act under Title V—Maternal and Child Health Services, Services for Crippled Children, and Child Welfare Services.Footnote ⁶⁸ Though elements of all three services had been taken up in the states by Sheppard-Towner agents, the template of Sheppard-Towner was integrated into Part 1, Maternal and Child Health Services. New federal matching grants revived the old state bureaus of maternal and child health that Sheppard-Towner funding had seeded. With the administrative apparatus of Social Security Act programs overlaid onto the old Sheppard-Towner ones, it was easier for states to get Maternal and Child Health Services off the ground than any other program under Title V. As soon as federal funds were made available in 1936, forty-seven states, Alaska, the District of Columbia, and Hawaii applied and received them. This compared to only thirty-six states that applied for funding for Services to Crippled Children and thirty-three that applied for Child Welfare Services that same year because these programs had not begun under Sheppard-Towner.Footnote ⁶⁹

The Children’s Bureau did not get as much control over maternal and child welfare policy as they had hoped for under the SSA. Most notably, they did not get to administer mothers’ pensions under Aid to Dependent Children. However, because of the state infrastructure they had set up under Sheppard-Towner, they were again put in charge of maternal and infant health programs under Title V. The Bureau was clear from the beginning that its goal with the SSA was primarily to reinstate its Sheppard-Towner agenda through the state bureaus. For example, when an Ohio woman inquired about the provisions of the new bill, a Bureau employee just sent her a report on Sheppard-Towner activities.Footnote ⁷⁰ Similarly, a publicity training manual for the Bureau taught staff to explicitly connect their work under the SSA with Sheppard-Towner. “Provisions of the Social Security Act continue and broaden the work already established in many States under the Sheppard-Towner Act,” the manual instructed.Footnote ⁷¹

When Senator Morris Sheppard (author of the original law) wrote to the Bureau to ask how the new law would continue the work of the old, Katherine Lenroot explained that Title V of the SSA was written “for the same purposes as were contemplated by the earlier Sheppard-Towner Act, and may be regarded as fully meeting the needs which the Sheppard-Towner Act was designed to serve.” She de-emphasized the distinctions between the two laws to Senator Sheppard, mentioning only that the new bill allowed states to work with children of all ages and would focus resources in rural and economically distressed areas.Footnote ⁷²

Title V of the SSA very explicitly stated that “the Children’s Bureau shall make such studies and investigations as will promote the efficient administration of this title.”Footnote ⁷³ It also clearly outlined that funding approval was contingent on states making “such reports, in such form and containing such information, as the Secretary of Labor may from time to time require, and comply with such provisions as he may from time to time find necessary to assure the correctness and verification of such reports.”Footnote ⁷⁴ By plainly laying out the federal agency’s prerogative to demand statistics in the form that they desired, the Children’s Bureau effectively compelled the states to comply with federal standards of data collection. It is in this legislative choice that we see the Bureau and Congress shift authority over matching grants from the state to the federal level. Henceforth, states would be required to be not only transparent in their use of federal funds but also uniform in their reporting.

Though it ran counter to their instincts and agenda, Children’s Bureau staff allowed a great deal of programmatic and administrative latitude to state directors in their Sheppard-Towner programs as they tried to maneuver within the politics of American federalism in the 1920s. Bureau employees were surprised to find that state directors had at best accounted for their programs unevenly and at worst had not accounted for them at all. Their inability to create national vital statistics and state program statistics not only frustrated them internally, but it also undermined their ability to defend their use of federal dollars to Congress. As we have seen, they took this lesson to heart and allocated themselves more administrative control over state-level accounting in subsequent legislation.

As Margot Canaday correctly asserts, the state must “puzzle” before it acts.Footnote ⁷⁵ The administrative history of the Sheppard-Towner Act shows that modern governments have to learn to make their bureaucratic apparatus and citizenry legible.Footnote ⁷⁶ Moreover, the politics of federalism made this process especially fraught in the early twentieth century. Through the defunding of the Sheppard-Towner Act and the programs’ subsequent rehabilitation under the Social Security Act, we see the progression of a federal bureau acting, then puzzling, and then acting with more precision. When we examine the problem of data collection in the evolving relationship between the Children’s Bureau and state bureau directors, we see one specific way that the politics of federalism constrained and even undermined federal authority in the early twentieth century. By understanding the particular role that statistics played in changing the power dynamic between federal and state bureaus, we can begin to understand how one federal agency developed its ability to “see like a state.”Footnote ⁷⁷