Labor Secretary Frances Perkins Reorganizes Her Department’s Immigration Enforcement Functions, 1933–1940: “Going against the Grain”

Abstract

Labor Secretary Frances Perkins championed liberal immigration policies between 1933 and 1940. Some efforts were successful, but most were not due to political, economic, and social constraints on immigration policy making, especially in Congress. Yet, she reorganized the enforcement functions of her department when she created the Immigration and Naturalization Service. Narratives abound about the period, though few delve into this reorganization. In this article, I share an analytical framework that I developed, “policy innovation through bureaucratic reorganization,” to explain how Perkins temporarily eased the debarments, as well as deportations, of newcomers by adjusting agency resources, including staffing, budget, and infrastructure. I describe how she responded to pressures from immigration restrictionists by tightening these functions. My narrative adds to the literature on immigration policy history, which has not fully appreciated the role of bureaucratic reorganization. This research bolsters the perspective in political control theory that bureaucratic structure merits as much attention as does legislation as a tool for control.

In 1939 Congress considered a resolution to impeach Labor Secretary Frances Perkins, who was the first woman to serve in a presidential cabinet and whose ideas—from unemployment insurance to Social Security—seeded many of President Franklin D. Roosevelt’s (FDR) New Deal programs. The charges against her and two department officials were, in part, that they “fail[ed] … neglect[ed], and refus[ed] to enforce the immigration laws of the United States.” ¹ The accusations were stunning because the Immigration Act of 1924 had made the immigration laws, to that point, the most restrictive in American history. Although the literature in immigration policy history recognizes the liberal policies she championed—some were realized, others were not—less is known about how Perkins reorganized her department’s immigration enforcement functions to ease the execution of those laws between 1933 and 1940.

This literature tends to equate policy change with legislation and has not elucidated the significance of Roosevelt and his labor secretary’s creation of the Immigration and Naturalization Service (INS or Service) in 1933 or her use of it until 1940. On one hand, Aristide Zolberg, who provides one of the leading narratives, contends that the immigration system was unalterable: it could not be used to alleviate the Jewish plight because of the strict immigration laws then in effect. ² On the other hand, Daniel Tichenor does consider how bureaucracies implement policies but argues that these laws prevented Perkins from increasing the admissions of newcomers. ³ Yet, others recognize the secretary’s policy advocacy efforts to facilitate admissions, such as by easing the issuance of visas, by confronting the State Department. ⁴ Although recognizing the State Department’s influence over immigration policy, Barbara McDonald Stewart nonetheless credits Perkins with the admission of refugee children. ⁵ And, even though Bat-Ami Zucker notes that Perkins needed to reorganize her department’s immigration-related functions and Rachel Brenner Graham writes that the secretary had power over immigration with the Service, neither researcher explains the structural significance of the INS in this period. ⁶ This observation points to the need for students of history and political science to look at what some might consider the same thing in new ways to see what we could learn.

In fact, Mae Ngai and Thomas Monroe Pitkin acknowledge that Perkins made administrative changes that reduced deportations. ⁷ The immigration policy history literature shows that, in the past, bureaucrats took similar actions to relax border control to permit migrants to toil for American businesses. ⁸ Kitty Calavita demonstrates how resourceful immigration officials allowed migration by engaging in “informal lawmaking.” ⁹ At the same time, immigration bureaucrats were just as inventive in devising ways to expel newcomers, including by pushing for their “voluntary departure.” ¹⁰ Also, immigration officials limited the mobility of laborers from Mexico to thwart their obtaining better wages and working conditions. ¹¹ S. Deborah Kang explains that Perkins encountered both compliance and resistance among staff regarding deportation policy. ¹² A better understanding of her influence over INS’ immigration enforcement functions—the debarment of new arrivals and the deportation of the undocumented—will show the opportunities and limitations of Secretary Perkins’s systemic actions to ease immigration restriction.

I use political control theory to examine Secretary Perkins’s efforts of going against the grain of immigration restriction. The theory explains how the power of Congress and the president, as well as lobby groups, shape policy implementation in public agencies. This scholarship considers the legislature to be more influential over agencies than the president is. In their seminal work, Mathew McCubbins, Roger Noll, and Barry Weingast assert that lawmakers push bureaucracies to adhere to their favored policies primarily by mandating administrative procedures. ¹³ In this vein, David Epstein and Sharyn O’Halloran contend that legislators enact laws that are “detailed” or “vague,” depending on their desire to dictate or delegate the execution of laws, respectively. ¹⁴ Detailed legislation can prevent policy drift: bureaucrats deviating from the policy preferences of politicians and the coalitions they represent. ¹⁵ When Congress combines procedure and structure to influence the execution of laws in government agencies, B. Dan Wood and John Bohte contend that “bureaucracies … are more difficult to change.” Their attributes can include “tightly bound discretion, clear mandates, and structural autonomy [from politicians].” ¹⁶ These insights suggest that between 1933 and 1940 lawmakers could have expected the specificity of the strict immigration laws to be followed by the Labor Department.

Led by Terry Moe, political control theory has evolved to incorporate the use of various forms of structure at the president’s (and Congress’) disposal. He argues that the structuring of public agencies bears directly on control of the policy implementation process. ¹⁷ In this “politics of structural choice,” Moe considers, in addition to Congress and the president, the roles of interest groups and bureaucratic officials in institutional change. He contends that interest groups drive such change. With respect to structure, researchers assert that features such as political appointments, budget, agency leadership, decentralization, and professional training can affect political control. ¹⁸ More recently, Stuart Shapiro shows that a feature of agency structure, the independence of analysts within an agency, influences procedural rules that are directed at maintaining the policy preferences of political principals. Equally important, he reminds us that bureaucratic structure can affect an agency’s outputs. ¹⁹ As far as the president, Moe and Scott Wilson contend that the executive can exert considerable power over the administrative state through the use of structure, such as centralizing authority. ²⁰ Nonetheless, William Howell and David Lewis demonstrate that presidential powers have limitations: presidents can create cabinet-level agencies to make sure policies are implemented according to their priorities when “Congress is relatively weak.” ²¹ Yet, Lewis also shows that legislators can hinder the president’s management efforts. ²² In sum, the theory helps to assess Perkins’s effect on debarments and deportations, but it needs coherence for application. The policy implementation process does not necessarily follow a straight line. Implementation can include twists and turns, like the contradictions in immigration policy making between rejecting and accepting newcomers.

Below I apply an analytical framework that I developed, “policy innovation through bureaucratic reorganization,” with the goals of contributing to political control theory and using this framework to reconsider accounts of Perkins’s leadership on immigration policy. ²³ My framework—consists of two phases: creation and evolution. ²⁴ The framework relies on the contention that officeholders can create complex bureaucracies to circumvent existing laws through policy implementation activities. They do so in part because the laws themselves may be difficult to change. Below I focus on the second phase, evolution, to describe the exact ways that elected officials can influence the execution of laws in government agencies. Evolution involves three steps. First, in a new bureau, officeholders engage in resource adjustment to relax the rigid implementation of laws. They add “liberal bureaucrats” who make administrative decisions to loosen operational rules. These officials or their agents reduce a bureau’s resources including its staffing, funding, and infrastructure. This step results in policy innovation: the agency’s administrative results (i.e., policy outputs) are liberalized.

Second, officeholders participate in coalition management . The structure of the bureau changes the benefit flows for competing interest groups. Officeholders and their agents use the liberalized results to cater to previously marginalized groups. To shift benefits to interests that oppose liberal immigration, officeholders can again use resource adjustment, expanding resources toward the implementation of existing (strict) laws, including through “restrictive bureaucrats.” Third, the president or Congress undertakes system redesign . The agency is reorganized because it produces disparate impacts; although some people benefit from its administrative results, the agency generates political, economic, and/or social aftereffects that harm others. ²⁵

I proceed in the order of these three steps. I begin by explaining the ways Perkins redesigned her department’s immigration functions as a structural solution to the stringent immigration laws in operation given the political, economic, and social contexts. Commencing in 1933, the secretary undertook several strategic actions including centralizing her authority over these functions through the inception of the INS. I use agency data to demonstrate that her efforts contributed to policy outputs of more newcomers admitted to and fewer of them deported from the United States. Next, I describe how she responded to immigration advocacy groups that had been marginalized by immigration officials. She empowered the advocates as they proposed reforms. They also benefitted from these outputs. However, Perkins also responded to pressure from restrictionists. Between 1936 and 1938, INS functions were tightened following criticism by Congress and the State Department. ²⁶ Also, as Maddalena Marinari makes clear, advocacy groups of Jews and Italians were not homogenous; they included pragmatists who accepted some immigration restriction to make gains on legislation centered on family unity. ²⁷ Last, Congress considered the INS a risk to national security, so in 1940, at Roosevelt’s request, lawmakers moved it from the Labor Department to the Justice Department.

RECONFIGURATION OF LABOR’S IMMIGRATION FUNCTIONS

Between 1933 and 1940 the economic, social, and political contexts of the United States largely constrained any action to ease immigration. The Great Depression promoted the notion that newcomers competed with Americans for jobs. ²⁸ The demand for work had decreased and the supply of unemployed workers increased. For its part, American society included strong anti-Semitic sentiments. ²⁹ Politically, legislative changes were blocked in Congress by the southern wing of the Democratic Party. ³⁰ To complicate the US immigration system further, the Immigration Act of 1924 made the State Department responsible for screening visa seekers abroad. The Act also limited the number of immigrants by establishing quotas. ³¹ The Labor Department continued to operate the part of the immigration system within the United States: it was responsible for screening and debarring new arrivals as well as deporting those who were undocumented. In 1933 its Bureau of Immigration admitted 23,068 newcomers; this number was the lowest since 1831, when 22,633 immigrants landed on American shores. The Bureau also deported 19,865 newcomers in 1933, at that point the most in history. ³² These data show that the bureaucracy Labor Secretary Perkins inherited that year was undeniably strict. Despite the barriers to immigration reform and the strictures of the Bureau, as secretary she had administrative power over the immigration functions under her purview.

In June 1933, FDR signed Executive Order 6144, with input from Perkins, which merged the bureaus of Immigration and Naturalization into the INS. ³³ George Martin notes that earlier she had informed the president that the Immigration Bureau was in need of reform. She was unsure about how to proceed, though she could not be indecisive. ³⁴ Her rationale for this structural innovation included a blunt assessment of the bureau as “disorderly; uncontrollable; unlawful [and] irregular.” She had deduced that these problems were due to a lack of administrative control. She found that too many departmental officials had their hands in immigration. Their numbers, along with the power they wielded, made them susceptible to influence by politicians, other public officers, and native-born Americans. ³⁵ Additionally, the secretary determined that authority within the system was too diffuse: there “were all of these districts” where decisions were made in the field. ³⁶ So, she reduced the districts by almost two-thirds, from 58 to 22. ³⁷ Much like Moe asserts, by resorting to reorganization, she achieved centralized control of the immigration functions in her department. ³⁸

Then the secretary homed in on personnel through addition and subtraction. She added a notable “liberal bureaucrat,” as she championed the president’s appointment of Daniel MacCormack as INS commissioner. ³⁹ Among other experiences, he served as a captain in the US Army during World War I. ⁴⁰ Perkins considered him to have “great administrative ability.” She was also impressed with his appreciation for the “dignity of each human being.” ⁴¹ This perspective aligned with her own views on immigration: “each one of these fellows who comes in as an immigrant, together with his wife and children, are human beings.” ⁴² Thus, the secretary and the commissioner were simpatico. MacCormack’s appointment served to advance their goals because, as Jennifer Nou asserts, agency heads “possess substantial discretion to impose internal structures and processes to further their interests.” ⁴³ In their first year, he subtracted 571 staff from the Service. He conducted an agency-wide assessment of employees to decide who stayed. One of the first groups of employees to be let go was the department veterans, “men with 30 years of service,” ⁴⁴ as arguably they were set in their ways and would resist implementing the secretary’s policies given their long tenures. As Moe notes, bureaucrats have career and institutional interests that might conflict with an agency’s mission. ⁴⁵ The assessment process—which included the US Civil Service Commission to safeguard the “merit system”—also affected other groups: clerical and nontechnical employees, as well as technical staff such as immigration inspectors, patrol inspectors, and naturalization examiners. MacCormack made no bones about “sift[ing] out the deadwood” throughout the agency. ⁴⁶

For the many employees who remained, the commissioner implemented a training program to mold their mind-sets. He sensitized the workforce to the struggles of immigrants by coaching them to “strive for that most difficult ideal—technical accuracy informed by justice and by humanity.” ⁴⁷ This standard was a dramatic change in the modus operandi of the personnel’s exercise of exclusion and expulsion powers. With respect to the former, a 1931 study found that inspectors “stretched” the limits of the law. For example, the likely-to-become-a-public-charge clause was a catchall when “officers think the alien ought not … enter, but the facts do not come within any specific requirement of the statutes.” ⁴⁸ But the training initiative instructed them to be more generous. Immigration inspectors at seaport and land border stations were required to be “courteous and dignified.” They were taught to be like a “welcoming committee.” ⁴⁹ Should a newcomer be detained for more questioning, inspectors were schooled to apologize for taking such action. ⁵⁰ Regarding the latter, a 1931 presidential commission had concluded immigrants were expelled even though “further development of the facts or proper construction of the law would have shown their right to remain.” ⁵¹ Yet the training regimen instructed staff to use “competent” interpreters because many newcomers did “not understand or speak the English language.” Given the adversarial nature of deportation cases, inspectors were forbidden from threatening or cajoling people to make incriminating statements. ⁵² As a precondition for exclusion or expulsion, MacCormack instructed his agents to develop the “facts both for and against the interests of the Government and of the alien concerned.” ⁵³

Ngai observes that the training program was an effort to “raise the professional level of INS personnel” ⁵⁴; however, this program furthered the liberal execution of policies and procedures. Marc Allen Eisner and Kenneth Meier show that organizational factors like professional training and expertise affect policy implementation even if they are not inventions of the president or Congress. ⁵⁵ With respect to the Service, the training program was a structural device that affected the agency’s exclusion function: during the inspection process, staff had authority to debar new arrivals. Previously, immigration inspectors manipulated the public charge clause to exclude newcomers when they were otherwise eligible. During the period in focus, the immigration laws were a constant, as was Regulation 3(J)(1), which provided “no hard and fast rule … as to the amount of money an alien should have.” ⁵⁶ Still, the training curriculum is instructive, as it was a critical change in how inspectors applied this clause. In the module on this process, there was a mechanism that permitted the “temporary detention” of a newcomer who was eligible for admission but for “possession of inadequate funds to reach the inland destination.” ⁵⁷ Immigrants were allowed to raise the funds they needed, albeit in detention, to gain admission. The connection between training and admissions is further supported by INS data: from 1934 to 1940 those denied entry based on this clause decreased by 18 percent, from 1,584 to 1,296. ⁵⁸ Also, the number of foreign nationals who entered carrying less than US$50 increased by 157 percent, from 6,678 to 17,137. ⁵⁹

Secretary Perkins used her centralized control over INS’ immigration and naturalization functions to facilitate admissions by having these functions complement each other. Like Moe and Wilson argue, a “centralization strategy” helps the president and, by extension, his political appointees to enforce managerial rules to influence bureaucratic behavior. ⁶⁰ Between 1933 and 1940, the Service processed naturalization applications that doubled the number of Americans, from 113,363 to 235,260. This increase was influenced by the secretary and the commissioner as they ordered naturalization examiners to cease opposing applicants for insignificant errors. ⁶¹ These new citizens, and those who had become part of the body politic before 1933, were poised to file applications for Rule 25(A) permits—authorized by Section 9 of the Immigration Act of 1924—so their family members could emigrate from the Third Reich. These permits were granted by the INS, not the State Department, and authorized consular officials to issue “nonquota” or “preference” visas to the kin of Americans. ⁶² The Service’s approvals of permits grew from 10,254 in 1933 to a yearly average of 15,046 between 1934 and 1940. ⁶³ This trend is corroborated by archival records, namely from the Honolulu district office, which experienced a doubling of Rule 25(A) requests. That office noted they were filed “mostly from persons of Jewish extraction desirous of assisting relatives to leave certain European countries.” ⁶⁴ Such efforts contributed to the inflow from the Reich as German-Jewish arrivals increased from 1,786 to 19,880 between 1934 and 1940. ⁶⁵ Surely these numbers were buoyed in 1937 by FDR’s decision to relax quota restrictions. ⁶⁶

In addition, the Service eased the entries of discrete numbers of migrant workers at the Southern and Northern borders by its administration of Rule 3(Q). It permitted the INS to issue identification cards to foreign nationals, from Mexico and Canada, who frequently crossed the borders for business or pleasure. ⁶⁷ The cards had allowed entry for employment until 1925. ⁶⁸ This prohibition continued in 1934, so only immigrants who permanently resided stateside could hold jobs. ⁶⁹ However, from 1935 to 1940 an annual average of 3,312 Mexican and Canadian laborers crossed the borders “daily or at least four times a week … with or without border crossing cards.” ⁷⁰ This development seems attributable to the secretary and staff. For her part, she opined that Americans preferred not to engage in manual labor ⁷¹ and that, historically, immigration provided laborers to America. ⁷² Her perspective aligned with another of her duties as labor secretary: the Employment Service “recruit[ed], distribut[ed], and direct[ed] seasonal farm workers.” It assessed the needs of farmers and moved laborers based on demand, ⁷³ which increased as the economy improved. In 1935 the Treasury Department determined that farm income grew; there were “gains of more than seasonal proportions in all regions.” ⁷⁴ As far as immigration staff, Patrick Ettinger points out that in the mid-1920s officials had agreed with growers to legalize their workers and that by 1930 officials had devised a mechanism to control Mexican migration to respond to growers. ⁷⁵ Contrary to criticisms of political control theory—as overemphasizing external control of bureaucracies while underappreciating internal control of them ⁷⁶—the extraofficial entries of migrants from 1935 to 1940 suggest there were shared interests and control between political appointees and staff.

Moreover, Perkins and MacCormack reformed the deportation process. They made an important change by prohibiting staff to make warrantless arrests of newcomers. The presidential commission had recommended that “an independent board” should issue arrest warrants. ⁷⁷ Staff could not be trusted with the responsibility to make arrest decisions. S. Deborah Kang notes the commissioner considered warrantless arrests unconstitutional; a warrant was a procedural safeguard for people. ⁷⁸ The secretary, too, abhorred such arrests and the subsequent detention of them for questioning. ⁷⁹ Warrantless arrests were “ordered discontinued” because warrants by law could only be issued with her approval; staff could telegraphically request them to expedite an arrest. ⁸⁰ These administrative actions were—as Jeffrey Pressman and Aaron Wildavsky demonstrate in their important study on policy implementation—“decision point[s]” that required “clearance” and produced “delays.” ⁸¹ In 1934 the commissioner reported that this prohibition meant “the disappearance of many suspects, particularly on the Mexican border.” He calculated that “2,600 persons … escaped who under the old system would have been held and deported.” ⁸² By 1940 the number of deportations was reduced from 19,865 to 6,954, or 65 percent, since 1933. Other agency activities apparently followed suit: repatriations decreased from 1,645 to 1,475, or 10 percent, and voluntary departures decreased from 10,347 to 8,594, or 17 percent. ⁸³ But this activity was not reduced as much as were deportations. Adam Goodman shows that in 1927 immigration staff were instructed to rely on voluntary returns as a means to avoid spending for deportations; these returns were faster than expulsions that necessitated 10 steps. ⁸⁴ The sizeable outputs for voluntary departures suggest that political control had its limits in the INS.

The secretary also curtailed deportations through budget maneuvers, much like B. Dan Wood and Richard Waterman assert about using agency funding as a method to affect administrative results. ⁸⁵ She eliminated the Service’s deportation squad, thereby reducing the resources that were used to expel immigrants. On paper, the squad was known as “Section 24,” named after a provision in the immigration laws that banned contracting foreigners to work in the United States. ⁸⁶ In practice, the unit far exceeded the law. Section 24 had been established in President Herbert Hoover’s administration, and its officers indiscriminately raided places where newcomers assembled. ⁸⁷ Perkins was well aware that the unit targeted labor leaders, who were of “foreign extraction” and labeled as “alien agitators,” to “quell labor troubles.” She further knew that “immigrants were shaken down by … this squad.” ⁸⁸ Equally troublesome, it arbitrarily sought mass expulsions. Places where foreigners congregated, such as lodging houses, were raided. In one case, 610 people were rounded up in Detroit. In another, “the entire Chinese section of Chicago” was stormed. ⁸⁹ Her challenge was not whether but how to rid her department of Section 24. The officers were tenured: they held “civil service status” and could not be dismissed outright. ⁹⁰ In response, Perkins let the unit’s funding—a $200,000 congressional appropriation—run out. Then she terminated its 53 officers because of insufficient funds. ⁹¹

More broadly, INS expenditures were reduced in 1934 and then incrementally increased each year until 1940 (Figure 1 immediately below). Expenditures were never higher than in 1933. Therefore, I use 1933 as a baseline for comparison with other years because it is the height of immigration restriction during this period: as previously noted, 1933 marked the historic low of admissions and a high mark for expulsions. Arguably, the agency needed to maintain spending at $10.4 million to achieve this level of restriction. As seen in Figure 1, in 1934 expenses totaled $8.5 million, a reduction of $1.9 million from the previous year. The US budget director had ordered the reduction. ⁹² Yet, the INS generated revenue through the collection of naturalization fees and head taxes, totaling $2.8 million in 1933. ⁹³ Conceivably, this revenue could have been used to offset the reduction in budgetary allocations from other sources, and maintained funding for the operation of the Section 24 squad. Instead, MacCormack noted that the spending reduction forced him to eliminate the personnel that executed the agency’s debarment and deportation functions. ⁹⁴ Indeed, Figure 1 shows that agency employment was reduced in 1934 from 4,105 to 3,524. According to one agency official, the workforce was “below the minimum required for the efficient conduct of business.” ⁹⁵ After 1934, as Figure 1 illustrates, the workforce mostly increased throughout the period. Generally speaking, this trend matches the increase in the admissions of immigrants and visitors, shown in Figure 2 (directly below). Also, Figure 3 shows that net immigration increased in the later years. (This figure appears further below.) There was demand on the INS. Yet spending never matched the level of expenditures reached in 1933, suggesting that immigration restriction was not the same priority for Perkins and MacCormack as it had been for their predecessors.

Figure 1. Expenditures (left scale) and Employees. ⁹⁶ (right scale)

Source: Department of Labor.

Figure 2. Admissions (left scale) and Debarments. ⁹⁷ (right scale)

Source: Department of Justice.

Figure 3. Net Immigration. ¹¹²

Source: Department of Justice.

The management of the Service’s infrastructure—the buildings, offices, and facilities where bureaucrats performed their duties—also reflected a disposition to assist immigration. Broadly speaking, I consider infrastructure included in McCubbins et al.’s definition of bureaucratic structure, which is “the allocation of resources and decisional authority among agencies and within an agency.” ⁹⁸ Thus, INS’ investment in its workspaces was a form of resource allocation. In them staff made critical decisions involving the inspection and detention of newcomers. This infrastructure supported 212 or so sites: 75 at seaports and 137 at ports of entry (in 1939). ⁹⁹ Figure 1 demonstrates that spending on immigration stations mostly decreased during the period, from $283,259 to $26,706. In 1934 the Service made one significant investment, dedicating $1.2 of $1.4 million from a Public Works Administration allotment to make capital improvements on Ellis Island. ¹⁰⁰ Pitkin notes that it had become, in effect, a detention center under Hoover, and Perkins wished to replace the island with another location. ¹⁰¹ Between 1937 and 1940, nonetheless, 611,910 immigrants and visitors from a total of 961,700 were inspected there. ¹⁰² Also, this investment responded to the Ellis Island Committee—an advisory board the secretary appointed—that recommended changes to the buildings and grounds. These enhancements were mainly programmatic, such as creating bigger rooms for social workers who advocated for newcomers at exclusion and expulsion hearings. ¹⁰³ The use of these funds evidences a priority to facilitate arrivals, not to detain and deport them.

The same can be said of many of INS’ border facilities because the secretary shared control over them with the Treasury Department. It was the lead in planning for the construction of new border inspection stations during the 1920s. The government had sought to tighten its borders to collect revenue—the Immigration Act of 1917 continued the requirement for immigrants to pay head taxes. Also, the Quota laws compelled the government to scrutinize the entries of “determined” newcomers who had failed to obtain visas. ¹⁰⁴ A 1928 report by the Treasury and Labor departments determined that the stations, to that point, were insufficient: inspections occurred in the open air, there was a lack of office space to thoroughly examine new arrivals, and locations were remote. ¹⁰⁵ From 1933 to 1940, these posts were heavily trafficked, as short-stay guests walked to America for business or pleasure, numbering 25.7 million annually. ¹⁰⁶ In 1934 the INS assessed its posts as operationally deficient. As far as debarment, border crossers were subject to “cursory inspection.” Many were not duly “manifested” or “statistically recorded.” In terms of deportation, they could not be thoroughly questioned because “it [was] a physical impossibility to detain” so many. ¹⁰⁷ During this period, 22 border stations were erected, plus 15 had been built between 1930 and 1932. ¹⁰⁸ But, they did not permit the INS to fully enforce its functions. The posts were shared between officers of the INS and the Treasury’s Customs Service, so spaces were limited even though their designs contemplated dual purposes. Also, only seven sites included areas for detention and the administration of hearings. ¹⁰⁹ The limitations of these border stations and the Labor Department’s shared control over them contributed to porous borders.

Figures 2 and 4 (immediately below) support the contention that during the period in focus the secretary’s reorganization contributed to reductions in the numbers of debarments and deportations. In this narrative, the INS is front and center, which considers Kenneth Meier and Laurence O’Toole, Jr.’s critique of political control research as overestimating the relationship between the political environment, including politicians, and agency outputs because “the bureaucracy itself is often left behind.” ¹¹⁰ I trace Perkins’s actions, which employ bureaucratic structure, to these immigration enforcement functions. The connection between her use of structural tools, such as the appointment of MacCormack, and the debarment and deportation functions is supported by political control theory. For instance, Wood and Waterman demonstrate that a political appointment “is very important” and “agency outputs shifted immediately after a change in agency leadership.” ¹¹¹ As to debarments in Figure 2, they decreased by 4 percent, from 5,527 to 5,300. Inspectors did not exclude new arrivals with the same frequency as in 1933. The door was more ajar at seaport and land border stations under her watch than during Hoover’s administration. With respect to deportations in Figure 4, as previously noted, they sharply declined in 1934. The INS curbed the brisk business of expelling immigrants. To be clear, these results in the context of the Holocaust are modest. Still, more newcomers were helped by the Service due to Perkins’s reorganization.

Figure 4. Deportations, Voluntary Departures, and Repatriations. ¹¹³

Source: Department of Labor.

Note: Figures in color in the online version only.

THE BALANCING ACT OF INTERESTS

These figures, using the first step of my analytical framework, elucidate the ebbs, but the flows between 1936 and 1938 are not apparent without appreciating the second step—the management of interest groups. For their part, Zolberg and Tichenor see lobby groups as more influential than government officials in the policy-making process. Their narratives consider these pressures as pivotal for the passage of laws that expand and restrict immigration. ¹¹⁴ More recently, Marinari adds nuance, as she points out that Jewish and Italian advocacy groups, even as immigration policy liberalizers, did not agree on strategy in their communities. She, too, focuses on their push in the legislative process. ¹¹⁵ In political control theory, these perspectives align with Wood and Waterman, as well as Moe. The former contend that interest groups attempt to influence policy through, among other means, publicity and lobbying activities. The latter asserts that they drive structural changes to affect policy implementation. ¹¹⁶ Alternatively, Jonathan Macey argues that the structural design of an agency—namely, its focus on a single interest group or multiple groups—can help to mitigate their influence. ¹¹⁷ In line with this argument, below I describe how the secretary responded to immigration advocates and restrictionists.

As immigration advocates pressed FDR and the State Department for more visas for refugees fleeing Nazi Germany, Perkins responded by giving them power to assess the Labor Department and recommend reforms. The focus on the State Department was logical because it controlled the first checkpoint to emigrate to America and, as noted earlier, Southern Democrats blocked reforms in Congress. Also, the State Department was change resistant, as David Wyman shows, because consular officers erected “paper walls” against Jews by rigidly administering the visa process. ¹¹⁸ This department had a slew of restrictionists in key posts, as Alan M. Kraut, Richard Breitman, and Thomas W. Imhoof point out, and anti-Semitism was prevalent. ¹¹⁹ Jewish advocacy groups sought to restructure this process to facilitate immigration, much like Moe contends interest groups do to realize policy preferences. These interests pressed FDR, with little to no success, both inside and outside his administration. On the inside, besides Perkins there were other prominent liberalizers seeking to reform the State Department, such as Henry Morgenthau, Jr. who lobbied it in 1933. ¹²⁰ On the outside, notable groups like the American Jewish Committee and the American Jewish Congress made the case to FDR. Breitman and Alan Lichtman assert that he was a “bystander” and did not provide the Jews aid until 1936. ¹²¹ The strategy to engage the presidency was, in line with Marinari, advanced by Jewish leaders and not embraced by other community members who opposed incrementalism. ¹²² Yet, Perkins immediately helped. With the advent of the INS, she formed the Ellis Island Committee, which was charged with a review of Labor’s immigration functions. She touted the 51-members as “nonpartisan,” though they were very much pro-immigrant, as a network of 33 social service agencies advised them. ¹²³ Advocates were, thus, poised to improve the second checkpoint of the US immigration system.

In response to these interests, the Service eased admissions and checked expulsions. In terms of the former, as previously discussed, Rule 25(A) passes allowed American families to, technically, bypass the State Department because their foreign relatives were exempt from the quotas or received preference visas. ¹²⁴ This outlet was further facilitated by the INS granting naturalizations as well as Jews tending to become US citizens, as Marinari asserts. ¹²⁵ Additionally, immigration officers debarred fewer arrivals based on the public charge rule; many more newcomers were admitted carrying less money. This development, which began in 1933, was exactly what liberalizers—inside and outside the Roosevelt administration—had lobbied so passionately for without success from the State Department until 1936. With respect to the latter, the elimination of the draconian Section 24 unit effectively offered newcomers some respite. ¹²⁶ Those who came as tourists, for instance, and overstayed their visas were less vulnerable to deportation. An equivalent action—the extension of visitors’ visas—was authorized by FDR five years later after Kristallnacht. ¹²⁷

Furthermore, the business lobby—agriculturalists and industrialists—was interested in unskilled and skilled laborers from Mexico and Canada, respectively. ¹²⁸ The lobby relied on them for labor and for a competitive advantage in the agricultural and industrial sectors. The Quota laws seemingly reduced the supply of workers from Europe, and the exodus of Southern Blacks to industrial centers in the North and South likely resulted in them replacing the Europeans. ¹²⁹ Nativists contended that these Black people were then succeeded by Mexicans in the agricultural sector. ¹³⁰ For their part, Northern industrialists relied on Canadians for their high-skilled labor. One estimate noted that 15,000 people commuted daily from Ontario to Detroit, ¹³¹ long the hub for automobile production. This reliance was exacerbated as the Great Depression precipitated the movement of Americans looking for work. Despite an oversupply of native workers, this lobby seemingly sought to exploit foreigners, especially Mexicanos, for profit maximization. They were coveted in states like Michigan and Texas because they were paid substandard wages; in California and Colorado, businesses kept production costs down by using Mexicans, who did not seek to unionize. ¹³²

The INS responded to these interests. The decision to prohibit warrantless arrests, as described above, meant that thousands of people were not arrested and expelled. Plus, the Section 24 unit was no more. These actions helped to maintain the itinerant workforce for businesses. To discretely increase it, the agency continued issuing Rule 3(Q) passes for workers to crisscross the borders. On top of these actions, its infrastructure, as explained before, was not optimal for inspecting and detaining border crossers. With respect to Mexicanos, the secretary was directly involved. For instance, she ended a feud between agriculturalists and Mexican farmhands. She did not mince words: on one hand, “the farmers … wanted the army to come out and force them back to work at the point of a gun,” and on the other, “these poor, idle, mistreated fellows were also perfect bums … they would get the best of you if they could.” To resolve the problem, she standardized wages and conditions, which stabilized this workforce for businesses. ¹³³ From 1933 to 1940, admissions from Mexico increased by 19 percent and deportations decreased by 50 percent; ¹³⁴ 7,258 Mexican day laborers also crossed the border. ¹³⁵ As far as Canadiens, a group of New York-based manufacturing companies in Niagara Falls retained their workers. ¹³⁶ The introduction of laborers was welcomed by the lumber industry as well, which routinely exploited migrants. In 1935, the agency reported that 15,000 Canadians trekked to northern New England to toil in lumber camps. ¹³⁷ During this period, admissions from Canada increased by 75 percent and expulsions decreased by 32 percent; ¹³⁸ 12,613 Canadian day laborers also walked back and forth across the border. ¹³⁹

The flows in Figures 2 and 4, particularly between 1936 and 1938, show that the Service tightened its functions, in part, to react to restrictionsts and, in part, because fissures emerged in the secretary’s control of the agency. First I discuss the pressures on the agency and then its reactions. Regarding the pressures, the INS carefully managed its statistics because the public worried about job competition from immigrants and restrictionists monitored their numbers. In 1935 the agency reported that “the problem of caring for the unemployed has not been aggravated by an influx of aliens to compete in the labor market.” The following year it represented that there was “a virtual balance … between total admissions and total departures.” ¹⁴⁰ However, in 1935 the American Coalition, an umbrella organization of 105 patriotic and civic associations, sent FDR a grievance of Perkins’s administration. The Coalition objected to the admission of refugee children from Germany as based on a loose legal interpretation and the elimination of the Section 24 squad. ¹⁴¹ These more than 100 groups, like Wood and Waterman contend about the means pressure groups use, lobbied the president to reverse Secretary Perkins. A related series of public critiques involved a deportation case that she oversaw, between 1934 and 1939, of immigrant labor leader Harry Bridges. He had led strikes by the longshoremen on the West Coast and was alleged to have been a communist; she could have deported him for subversive activities but delayed a decision to benefit from court judgments on a related case. ¹⁴² Kristen Downey notes that the Bridges matter earned the secretary criticisms by the public and in the press, as well as a congressional investigation and an impeachment resolution. ¹⁴³

In reaction, the INS adjusted resources to expand enforcement activities, though staff went further to push out more people. Regarding debarment, Figure 1 shows that expenditures and employees increased: starting in 1935, reaching an apex in 1936, and becoming static in 1937. Figure 2 depicts increased admissions during the period, except in 1940, and debarments generally following the same trend. A comparison of these numbers connotes similar trends among employees and debarments during the period, with the exception of 1939 when staffing increased and debarments decreased. This decrease might have been because staff was overwhelmed by admissions, which were the highest during this period, at 268,331.

In terms of deportation, in 1936 MacCormack modified the rule on staff’s power to arrest. On its face, the change was relatively minor, as it specified the administrators who could designate staff to execute warrants. But, in practice he reversed the prohibition on warrantless arrests from 1933 because officers were advised that they could make such arrests as long as they witnessed the individuals unlawfully entering America. ¹⁴⁴ Figure 4 shows that this revision coincided with increased deportations in 1936 and 1938, with a decrease in 1937. Otherwise, a comparison of Figures 1, 3, and 4 does not show similar trends for expenditures, employees, net immigration, and deportations, except in 1934 and 1936 when these four indicators decreased and increased, respectively. ¹⁴⁵ Figure 4 further shows that voluntary departures increased between 1936 and 1939 and repatriations sharply increased from 1937 to 1939. The increases suggest that with one less constraint, staff members intensified their use of these three enforcement tools. Also, Kelly Lytle Hernandez contends that MacCormack was unable to purge some of the “old-timers,” who worked on the borders, and they negatively swayed new recruits. ¹⁴⁶ Regarding deportations, Kang points out that some bureaucrats deviated from the agency’s reforms. ¹⁴⁷ As noted earlier, Goodman asserts that voluntary departures were a quick, informal way to expel people. ¹⁴⁸ Relatedly, the repatriation numbers are telling of staff’s harsh preferences; in 1937 the immigration laws were amended to make the repatriated ineligible to reapply for admission. This condition insinuates that staff used this tool to bar people from returning, as voluntary departures gave them the option to reenter. ¹⁴⁹ In comparison, as seen in this figure, the numbers for repatriations expanded.

INS TRANSFERRED TO JUSTICE

With the last step of my framework, I explain the transfer of the INS from the Labor Department to the Justice Department in 1940. Inside the agency, Secretary Perkins’s reform efforts were waning, as reflected by a report she had commissioned. In her autobiography, she claims to have repeatedly recommended that the agency be moved out of her department, which was “swamped” by the Service. ¹⁵⁰ Outside the agency, it was deemed a risk to national security. Lawmakers and Roosevelt considered the INS lax on enforcement and fretted about the prospect of America entering World War II. ¹⁵¹ Congress and the State Department pushed for this transfer. For political control theory, the Service’s demise provides an opportunity to assess the relative strengths of some of the structural tools from the literature, such as political appointments, as well as the possible lifespan of a such an agency. Moe theorizes that the politics of structural choice is a perpetual process, ¹⁵² yet the INS, as conceived, ended.

An internal review of the INS that she initiated in 1938 indicated that Perkins’s control had slipped. This review, which resulted in a report tendered to her days before the president announced the relocation of the agency in 1940, largely critiqued its deportation process. The committee that was conducting the study opined that centralized control over arrest warrants had exposed the process to staff manipulation. When staff members requested warrants, in many instances they did so via telegraph, which allowed them to obtain the warrants while providing little information to headquarters. ¹⁵³ As part of this concern, the committee recommended a reorganization of the agency, which included decentralizing control of this process. The committee further proposed the creation of an executive officer position at headquarters. Perkins implemented this proposal. ¹⁵⁴ In making it, the committee reasoned that coordination of agency activities previously undertaken by MacCormack should be performed by this officer. ¹⁵⁵ Yet, the report is silent about his successor, James Houghteling, who ran the Service from 1937—after MacCormack’s death—to 1940. In short, the report was also a reflection of the new commissioner’s leadership. ¹⁵⁶ Relatedly, Stewart makes a critical observation about the difference between them: “[t]he former [c]ommissioner had tended to do what he thought right if there was no law forbidding it. The new [c]ommissioner looked for a law to give permission first.” ¹⁵⁷ In other words, the incumbent was passive. All these circumstances intimate that the secretary’s control over the agency may well have been firmer under MacCormack. Plus, the report does not appreciate the discretion employed by staff, as noted above by Calavita, Ettinger, Goodman, Hernandez, and Kang, to capriciously enforce immigration rules.

Congress and the State Department undermined the secretary’s leadership of the Service. As referenced earlier, lawmakers investigated and sought to impeach her. In 1937 Representative Martin Dies (D-TX) led an inquiry on “un-American activities” and reported that “Nazi-Fa[s]cist groups” freely engaged in subversion. His committee concluded “that the failure of the Labor Department to carry out the laws with respect to deportation [was] a contributing factor.” ¹⁵⁸ Two years later, Representative J. Parnell Thomas (R-NJ) introduced House Resolution 67 to impeach her for, in part, failing to deport Bridges. Although this motion was rejected by the House ¹⁵⁹ and the Supreme Court subsequently vindicated her by determining that expulsion was unjustified, ¹⁶⁰ legislators second-guessed the deportation process: “the confidence of the people in the proper administration of the laws … has been profoundly shaken.” ¹⁶¹ Along this line of attack, and the previously mentioned Labor–State tensions, in 1940 Undersecretary Sumner Welles urged FDR to move the INS to Justice because Labor had a “duality of interest” between managing the labor market and immigration enforcement. ¹⁶² Although she had misgivings—the Justice Department was ultimately not the appropriate place for the Service because it dealt with “human affairs”—FDR presented Reorganization Plan No. V to Congress, which approved it in June 1940. ¹⁶³ He publicly promised “more effective control over aliens” and that INS would provide for the “nation’s safety.” ¹⁶⁴

CONCLUSION

In her statement to Congress against impeachment, Secretary Perkins justified easing immigration enforcement at Labor: “I have imposed restraints upon the arbitrary use of this power, and I have sought to build and maintain confidence in our institutions by proceeding in cases with scrupulous fairness.” ¹⁶⁵ According to DeLysa Burnier, the impeachment resolution, and other attacks she faced, marginalized Perkins. Burnier contends further that, in the discipline of public administration, the secretary was a “devalued subject” because by “profession and gender” she was a “settlement woman.” This discipline had focused on the “bureau men” of the Roosevelt administration, including their reorganization of the executive branch. Thus, my narrative “provid[es] counter-readings … to what appear to be settled knowledges and practices.” ¹⁶⁶ Her reconfiguration of the bureaus of Immigration and Naturalization between 1933 and 1940 receives little attention in the political control and immigration policy history literatures. Below, I discuss the insights that the use of the analytical framework of policy innovation through bureaucratic reorganization adds to these literatures, especially as I used the former to elucidate the latter.

With the first step of this framework (from the evolution phase), we can understand how Perkins adjusted the agency’s resources to ease and tighten debarments and deportations. Zolberg argues that prior institutional capabilities constrained policy makers. However, as explained above, through a combination of structural devices, she loosened the constraints and liberalized these activities. For example, the structural devices, such as political appointees, facilitated the processing of more Rule 25(A) permits to help Jews with American kin escape the Third Reich. The noted record from a district office confirms this development. By delving into the weeds of this bureaucracy—something political control theory has undervalued—we realize that the admission process was not practically “impenetrable,” as Tichenor asserts. ¹⁶⁷

Perkins employed the same devices for immigration restriction, except that funding and staffing did not have as much influence on deportations as they did for debarments. In particular, MacCormack reversed the decision prohibiting warrantless arrests, which not only resulted in increased deportations but also allowed staff to intensify voluntary departures and repatriations. We can infer that, after MacCormack’s passing, Houghteling—as a political appointee—did not maintain the centralized control Perkins had established. Contrary to the criticism that political control theory overvalues political influence, this case reveals that such power was affected by staff discretion, which Ettinger contends was developed before 1933. By tracing bureaucratic structure and policy outputs to political control, we can learn how this control can be used to quickly change the direction of outputs; Whitford argues output shifts can occur when Congress and the president compete for control. ¹⁶⁸ This research helps make sense of the seemingly contradictory and conflicting nature of immigration policy.

The second step of the framework, coalition management, elucidates the agency’s dynamic, yet limited, responses to lobby pressures for and against immigration. These responses have implications for political control theory, as Macey asserts that such pressures can be assuaged through bureaucratic structure as designed at an agency’s inception. ¹⁶⁹ As explained above, INS’ enforcement activities were eased in response to Jewish advocacy groups, industrialists, and agriculturalists, as more newcomers entered and remained in America. Conversely, the Service reacted to the more than 100 associations that complained to FDR about its laxity. These shifts in policy outputs fall outside Macey’s perspective because resources were adjusted after the agency’s commencement. Thus, outputs can be altered in real time, affording flexibility to political appointees to react to interest groups. But, as this case also shows, managing interest groups had limits. For Jews, the INS could not respond to those who clamored for comprehensive immigration policy reform. Also, after Kristallnacht the agency’s incremental efforts were insufficient. Indeed, other liberalizers in the Roosevelt administration like Morgenthau, Harold Ickes, and Eleanor Roosevelt advocated to relocate refugees to places outside the US mainland. ¹⁷⁰ With respect to Mexican laborers, the issuance of Rule 3(Q) passes continued the inequitable practice of making them an itinerant workforce. Hernandez contends that immigration officers controlled this workforce. ¹⁷¹ Even the pivot to restriction had repercussions, as staff redoubled its activities to push people out.

The third step of my framework, system redesign, helps us to appreciate the catalyst for the Service’s redesign and its lifespan. In terms of INS’ demise, the primary consideration for FDR and Congress was that, as WWII loomed, its immigration enforcement functions compromised national security. That was the disparate impact of easing immigration. By the numbers, Figure 3 depicts that net immigration increased by 28,875. On top of that, 19,871 day laborers entered between 1935 and 1940. ¹⁷² These data represent a fraction of the 132 million Americans in 1940. ¹⁷³ Still, this case makes clear that INS was a temporary structural solution to immigration restriction. The temporary nature of the INS merits further study. Moe contends that structural politics is perpetual. In contrast, the agency had a life of seven years in the Department of Labor. This implication requires more research. For instance, the only other time the bureaus of Immigration and Naturalization were combined was in 1906, and they were disbanded in 1913. Moreover, such research is important for political control theory, as he argues it has pushed structural politics aside. Moe urges scholars to “push against constraints … be open to abnormal ideas and think actively about shifting the theory onto new paths that might be more productive.” ¹⁷⁴

Speaking of out-of-the-box ideas, the timing of reconsidering Secretary Frances Perkins’s leadership on immigration policy is apropos. The pandemic we are living through has laid bare and exacerbated the inequities produced by our public institutions. The statement from her impeachment defense is instructive: Perkins wanted the bureaucracy to treat people fairly. Her reorganization of the immigration system in the United States was a way to do that. For a brief period, she made this system more democratically accountable to the minority, when immigration restrictionists in the majority in Congress opposed passing comprehensive legislation to protect foreigners from persecution. Therefore, Secretary Perkins’s efforts are an example of going against the grain to make public institutions more equitable.