Conquering Space Through Internal Improvements: Federal Nation-Building in 19Th-Century America

Conquering Space through Internal Improvements: Federal Nation-Building in 19th-Century America

October 2019

Word Count: 8,623

Abstract

Early American political leaders were tasked with sustaining a representative republic on a seemingly impossible scale. Their struggle to stave off political dissolution raises an important question for scholars of federalism. How can democratic governments integrate disparate political communities across a vast – and rapidly expanding – territory? We revisit the solution most-often proposed by contemporary political leaders: a nationally directed system of internal improvements. Using a dataset of 19th-century appropriations, we find that patterns in internal improvement funding are consistent with a nation-building strategy. Congressional districts at the fringes of the republic received disproportionate support from the federal government, even after accounting for political preferences, positions of legislative authority, and sub-national spending patterns. Our research stands in contrast to existing work on internal improvements – which is primarily interested in testing theories of distributive politics – and contributes to a diverse body of research on federalism, nation-building, congressional politics, and American political development.

From the earliest days of the American republic, the size of the nation presented a challenge to the viability of the nascent government. Direct democracy was understood to be unworkable across an expansive territory, and no country had successfully established a representative republic on such a scale. While the United States was undoubtedly rich in natural resources, the economic promise of America’s physical geography simultaneously imperiled the young nation’s tenuous political stability.

The Founders constructed the American Constitution to be “partly federal, and partly national” to allow representative government to thrive in a large, diverse place (Federalist 39). That carefully balanced structure required constant upkeep in the years following ratification. The health of federalism was so centrally important to the national government, in fact, that differing solutions advanced by Hamiltonian Nationalists and Jeffersonian Republicans would sketch the fault lines of

American political ideology for years to come. The problem of governing a vast political domain without the complete consolidation of federal power proved to be a thorny, recurring motif in the development of U.S. political institutions.

In other words, American political leaders had a nation-building problem. To survive, democratic regimes require secure territorial boundaries, adherence to the rule of law, and strong economic, social, and political integration.1 Governments that effectively connect citizens to one another and to their central political institutions thus succeed in nation-building. While greater integration promises stability, a failure to effectively construct the nation threatens dissolution.

Given the stakes, it is not surprising that America’s Founding Fathers eagerly engaged in early

1 “Nation-building” is a widely used and rarely defined term. We primarily think of nation-building as a particularly important form of integration. While “integration” is defined in various ways, each definition is consistent with what we believe is the common understanding of nation-building. Integration primarily invokes efforts to mix groups, to take separate parts and make them whole, and to coordinate processes in a complex system. In short, we view nation-building as an effort to take diverse constituencies and render that fragmented political community more complete. 1 projects of nation-building to overcome the very real threats of sectionalism, economic competition, and political chaos.

Nation-building is critically important to any regime, but America exacerbated the nation- building challenge through a constant impulse for expansion. As settlers moved west and America’s political leaders dreamt of an empire stretching to the Pacific coast, the challenges of governing a large republic under a mixed regime grew. In this sense, the central problem of nation-building for

America in the early 19th century was one of overcoming space. Connecting people to one another, uniting material interests behind an integrated national economy, and establishing the pre-requisites of an imagined community with responsive representation became paramount (Pears, 2017).

Few policy areas illuminate this challenge as clearly as the attempt to establish a bold, nationally sponsored system of infrastructural projects throughout the first century of American history. Internal improvements became, for a large group of 19th-century statesmen, a key tool in the fight to bring citizens closer to one another and the institutions of the federal government.

Between 1789 and 1882, Congress allocated hundreds of millions of dollars for infrastructure and development. These efforts, ranging from the construction of the national road to the dredging of massive harbors, would be the equivalent of a roughly $5.6 Billion investment by today’s standards.

In short, federal funding for internal improvements was among the most important and contentious issues in 19th-century American politics.

While these projects were almost universally recognized as politically and economically beneficial to the United States, fears of consolidated federal power drove ideological battles over whether and how the federal government should pay for improvements within state borders. As a result, systematic plans for nation-wide infrastructure development were never fully enacted, and piecemeal federal and state-level funding became the norm. Nevertheless, the allocation of federal funding for internal improvements provides a lens through which we can view American nation-

2 building efforts and reconsider existing research on federalism, congressional politics, and the development of political institutions.

Most political science research specific to internal improvements considers the allocation of federal funds from a distributive politics perspective. In part, this research reflects a growing interest among political scientists in leveraging 19th-century cases to empirically investigate modern theories of lawmaking (Jenkins and Gailmard 2018; Finocchiaro and Jenkins 2015). In this spirit, Wilson

(1986) finds that river and harbor legislation generally reflected a universalistic pattern of distribution. More broadly, Minicucci (2004) looks at the pre-war period of internal improvements and concludes that the introduction of national political parties in the 1830s disrupted the tools of distributive politics used before that period, forcing an adjustment in the politics of pork-barrel appropriations. Gordon and Simpson (2018) similarly employ an exhaustive dataset on internal improvement appropriations to make a compelling argument that internal improvements follow a familiar pattern of distributive politics, complete with universalistic logrolls, but this trend only emerges in the last quarter of the 19th century. In contrast to these studies, Finocchiaro (2015) detects a distinct majority party advantage after analyzing public building projects at the end of the century. Despite their differences, each of these studies is primarily interested in the distributive nature of internal improvement projects. This focus is somewhat surprising, as some of the most prominent political actors in American history plainly stated that internal improvements were designed as a critical component of nation-building and a means of connecting citizens to their central government.

We believe our understanding of Congress has been greatly enriched by the move to test contemporary theories of politics using historical data, but in contrast to that body of research, we do not approach these projects as an opportunity to better understand the roots of distributive politics. Instead, we evaluate whether appropriations for federal projects in the 1800s adhered to

3 political leaders’ avowed nation-building goals. More specifically, we advance a simple question: is the distribution of federal funds consistent with a nation-building perspective? In other words, is it possible lawmakers used internal improvements to connect citizens and conquer the space that separated Americans from their central government? In this sense, we revisit the logic most-often articulated by political leaders in the era to assess whether implementation followed rhetoric.

Following Minicucci (2001), we take seriously the possibility that nation-building – among other motivating factors – played a significant role in the designation of federal investments. If claims of union and nation-building were truthful, we should be able to identify the political communities that disproportionately benefited from these projects.

Investigating internal improvements from this perspective also allows us to link existing research on legislative affairs with empirical work on nation-building among comparative and

American political development scholars (Adler 2012), but political scientists are not the only group of scholars interested in the subject. Historians often describe internal improvements as an important case of failed nation-building. For example, Larson (2001) argues that Americans –

"tutored constantly on the dangers of big government and public power" (p.5) – squandered an enormous opportunity to implement a truly national plan and, along with Goodrich (1960), claims that state and local efforts dwarfed those of the federal government. With few exceptions (e.g.,

Malone 1998), historians argue that piecemeal internal improvements failed to deliver nation- building outcomes. Nevertheless, much of the careful research on this subject matter both assumes a primacy to the nation-building logic of internal improvements and declares these efforts an informative failure of American political history. Whatever the impact of the projects, we maintain that the apparent failure of internal improvements to overcome the challenges of sectionalism and economic competition should not be taken as an indication of their intent. On the surface, internal improvements certainly appear to demonstrate a policy failure. The bloody Civil War and the

4 notable absence of anything on the scale of Henry Clay’s proposed American System seem to provide a proof of concept. However, internal improvements may have both failed to pull the nation back from the brink of war and succeeded in improving the integration of disparate political communities. To account for this possibility, we need to grapple with an unobservable counterfactual: a 19th-century America without internal improvement spending. In other words, we do not know what the probability or timing of dissolution would have been without this piecemeal attempt at nation-building.

In this research, we want to know if the distribution of internal improvement appropriations systematically suggests a federal attempt to better integrate the new American republic. In the following sections, we provide an overview of the particular obstacles of space and geographic area to American nation-building. Towards this end, we highlight the proposed solutions present in 19th- century political theories, and we outline how key political leaders turned to internal improvements in an attempt to stabilize the large republic. We provide brief qualitative descriptions of ow various project types each contributed to the nation-building enterprise. Next, we test the nation-building perspective by evaluating the factors corresponding with increased investment in political communities across the country. Following Gordon and Simpson (2018), we execute a two-way fixed effects research design utilizing official government records of appropriations for public buildings, river and harbor improvements, forts, arsenals, armories, lighthouses, and a variety of other public works projects throughout much of the 19th century.

In short, we find evidence in support of the rhetoric put forward by proponents of the

American System. Congressional districts from states at the fringes of the republic appear to have received disproportionate internal improvement funding — even after accounting for various indicators of legislative influence and a record of state-level investments. These findings are consistent with the claim that lawmakers funded internal improvements projects to overcome the

5 challenges of distance and construct a truly national union, suggesting that a nation-building perspective played a significant role at this critical juncture in the continued development of the

American federal system.

Nation-building and the Need to “Conquer Space” in an Extended Republic

In early American history, problems of dissolution and nation-building were paramount.

Foreign invasion was a very real possibility – indeed a reality – and border disputes in the remote

North and West posed a constant dilemma. From outright political rebellion to pitched nullification crises, the national capacity to enforce the rule of law in distant corners of the union remained, even after the abandonment of the Articles of Confederation, something of an open question.

Sectionalism threatened to tear the nation apart and was by no means confined to the south, and while slavery unquestionably drove the nation’s most important internal conflicts, the Hartford

Convention made clear that regional loyalty posed a serious, national problem (Borneman, 2004).

Economic competition between regions and interests drove tariff battles and fueled precisely the kind of division the Founding Fathers feared most. In short, the results of the American experiment were knife-edged and preliminary at best, and the fragility of the nation in many ways stemmed from a lack of political, social, and economic integration.

Moreover, the particular nature of American geography vastly exacerbated these problems of nation-building. At the Constitutional Convention in Philadelphia in 1887, America’s Founding

Fathers returned again and again to a central problem: the sheer geographical size of the American confederation made it too difficult, maybe even impossible, to govern. While nationalists at the convention believed a single, energetic, republican government was the best way to secure the people’s liberty, they recognized that the fundamental features of republican governance, including representation and accountability, required commercial intercourse and effective communication –

6 which seemed to be undermined by the size of the republic. Put differently, “in the sprawling continental union nothing threatened the mutual interests of the citizens and their states like geographical isolation” (Larson, 2001, 3). Ultimately the Federalists and their intellectual successors believed they could overcome the problems of a large republic with well-placed programs of nation- building. Such projects, they argued, could ease the divisions that emerge within an expansive territory by cultivating the connectedness naturally present in smaller republics.

There were many tools of nation-building in this time period, and each aimed, in their own way, to overcome the challenges of disconnection and dissolution. In the early 19th century alone, congressional leaders used land grants, the postal system, trans-continental railroads, tariff policy, national party organizations, parades, songs, celebrations, civic education, newspapers, and many other policies in an attempt to connect citizens to one another and to the institutions of the federal government. In each case, nation-building efforts fought against the dissolution that comes with diversity and distance. Railroads and the postal system connected citizens and allowed for the spread of information across geographic boundaries. Civic education and celebrations reminded citizens of their shared interests and common pursuits. Land grants become tools of directed settlement that helped the American regime lay claim to lawless territories. While nation-building can and did come in many forms, we find three broadly defined mechanisms in the rhetoric of these 19th-century nation-builders.

First, connecting people to one another and to their representatives is key to constructing shared interests and facilitating communication. By developing transportation infrastructure and facilitating the dissemination of goods and information across the continent, the American government could create the conditions for nation-building that James Madison deemed most important (Madison 1815). Second, democratic citizens, particularly at the edges of a nascent and fragile regime, demand security. By clearly and actively defending borders and ensuring the safety of

7 territorial settlers, the federal government in Washington could reinforce the idea that the far reaches of the union were just as much a part of its political compact as the Atlantic communities that dominated its founding. Finally, beyond specifically protecting territorial boundaries, Congress could engage in nation-building by making its peaceful and operational presence known. As Publius argued in the Federalist Papers, a government with a physical presence in each state of the union would be far better able to win over the allegiance of its citizens than one that appeared to neglect the places where operations were less convenient. Transportation, security, and visibility each provide distinct benefits, but each mechanism assists in integrating the American political and economic community.

Projects of internal improvement stood out as the tool with the most promise to utilize all three of these nation-building mechanisms. Improvements aimed at transportation infrastructure helped to connect citizens, uniting their material interests and encouraging the interregional communication that in turn fosters healthy republican politics. Many of these projects improved military infrastructure, which demonstrated that the federal government protected states and citizens across an often-violent continent. From preserving international borders to enforcing the rule of law, improvements helped Congress provide the security that American citizens were promised by the federal compact. Finally, internal improvements demonstrated the utility of the union and the federal administrative state. Improvements to transportation infrastructure, the construction of new buildings, roads, and bridges, and the facilitation of local commerce all showed Americans in distant parts of the Union that their government was useful, present, and unitary.

Generations of America’s most notable statesmen supported this claim, as they routinely argued that internal improvements would be required to connect disparate citizens and facilitate the development of national interests. Alexander Hamilton, for example, provided a near-perfect expression of the ways that internal improvements might remind citizens of the utility of their union

8 in the face of the diminishing affections of distance. Writing in Federalist No. 17, he argued that human nature’s “affections are commonly weak in proportion to the distance or diffusiveness of the object” and that as a result, “the people of each state would be apt to feel a stronger bias towards their local government than towards the government of the Union.” As Madison argued, a republic could govern a large geographic territory only “if conditions that ease communication among the citizens are present” (Sheehan 2009, 98). Calhoun, too, diagnosed the problem of physical separation in the republic, but he also provided a clear programmatic remedy:

…what can be more important than a perfect unity in every part, in feelings and sentiments? And what can tend more powerfully to produce it than overcoming the effects of distance? …Whatever impedes the intercourse of the extremes with this, the centre of the republic, weakens the union… let us, then, bind the republic together with a perfect system of roads and canals. Let us conquer space (Calhoun 1888).

Clay’s American System, which Calhoun supported, “was intended to strengthen the bonds that tied the nation together into a single whole. It was intended to ensure the perpetuity of a united country”

(Remini 1991, 225). Speaking on the House floor in 1824, Clay asked: “is not the union best invigorated by an intimate, social, and commercial connexion between all the parts of the confederacy?” (ibid, 226). A republic covering such a large land area would have to develop a strong transportation infrastructure to support economic and military endeavors, but also to ensure the continued integration of national sentiments.

Support for federal funding for internal improvements was by no means universal, but it is important to note that even opponents of federal funding for internal improvements, like Andrew

Jackson, conceded the need for nation-building and the development of transportation infrastructure. In the case of the Jacksonians and many other skeptics of national power, federal funding for internal improvements amounted to a violation of constitutional restraints on congressional power. Anti-consolidationists frequently sought funding for internal improvements at the state level for this reason (Ellis 1987).

To our knowledge, however, scholars have not yet evaluated whether systematic patterns in the distribution of federal funds were consistent with these rhetorical calls for greater integration.

Based on this historical justification, we consider a nation-building hypothesis for the allocation of federal appropriations for internal improvement projects through 1882. If internal improvement projects were integral to a nation-building strategy, we expect to see that lawmakers disproportionately invested in those political communities with the greatest need for such nation- building efforts, namely the most disconnected corners of the republic.

Internal Improvements as Tools of Nation-building

Before turning to an analysis of the data, it may help to provide a more concrete illustration of the internal improvement projects advanced by prominent political leaders. We offer a description of federal funding for internal improvements in and around a single city, Portland,

Maine, drawing parallels to other projects across the country. While Portland provides a convenient and geographically interesting illustration of how these projects contribute to American nation- building, we also take care to incorporate examples from California, South Carolina, and several other states. These descriptions better identify the mechanisms by which these projects contributed to a broader goal of national integration.

As a visitor, the expansive harbor likely stood out as one of the first and most notable features of the city. Portland Harbor served as the state’s transportation center and chief economic port – particularly before the arrival of major roads or railroads connected Maine to the cities of

Boston and New York. Portland, and the city of Bath just to the north, were vital economic ports for the shipbuilding and lumber industries. Both citizens and trade goods arrived and departed from that harbor, and upkeep, including dredging and the maintenance of piers and embankments,

10 became crucial means of protecting the state’s social and economic lifeline to the rest of the nation.

Fortunately, Congress allocated significant funds for dredging and fortifying Portland Harbor between 1836 and 1882. In this respect, internal improvements in Maine were similar to those in San

Francisco Bay, Boston Harbor, and Galveston Harbor, which served to connect those cities to one another and to the broader network of coastal trade and communication.

Congressional investment in rivers, as well as harbors, in Maine allowed Washington and southern states to take advantage of Maine’s primary manufacturing industries and linked that distant state to the growing industrial economy of the mid-Atlantic. Improvements to the Penobscot and Kennebunk rivers proved to be essential. These rivers facilitated the transportation of lumber for the shipbuilding and paper industries from the far northern, British border of Maine to Portland, where it could, in turn, be loaded onto ships bound for cities across the union.

Here, too, Maine illustrates a more general trend in internal improvements. Across the country, riverways proved vital to America’s economic integration, connecting otherwise disparate regions. For example, Calhoun specifically highlighted the transportation potential of the Mississippi

River as “an arm of the sea” to be treated as an extension of the coastal fortifications so valuable to international trade that even strict constructionists would feel compelled to offer their support

(quoted in Blanchard 1894, 347). Beyond funding for projects along the Mississippi Delta and

Atlantic coast, evidence of funding for rivers can be found in states like Oregon, which was appropriated considerable sums of money to improve the Willamette and Columbia rivers on over thirty occasions.

In Maine, riverways proved vital aids in transportation, while in other parts of the country deltas and swamplands hindered communications. The congressionally funded road from Memphis to Little Rock provides an interesting example. Originally authorized as a military road in 1824, the east-west route across the Arkansas swamps eventually carried thousands of settlers headed west and

11 facilitated the forced removal of thousands of Choctaw and Creek tribe members in the 1830s. The road received congressional funding throughout the 1820s and 1830s as the Army Corps of

Engineers encountered repeated challenges in their attempts to construct bridges and appropriately drain the roadway (Longnecker 1985). Congress similarly approved money for the construction, extension, and survey of canals across the republic, including major projects in Virginia, Florida,

New Orleans, and Milwaukee. In many cases, canals connected smaller rivers to major transportation waterways like the Mississippi, and allowed goods to flow into and out of less connected areas.

Upon entering Portland Harbor by sea in the latter half of the 19th century, a visitor to

Maine would encounter two large coastal fortifications: Fort Scammel and Fort Gorges. Fort

Scammel was originally built on House Island in Portland Harbor in 1808 as an octagonal wooden blockhouse, but congressional funding, beginning in the 1840s and continuing through the Civil

War, allowed for significant refortification. Following the War of 1812, the Army Corps of

Engineers recommended the construction of Fort Gorges, in Casco Bay, to provide additional support for Fort Scammel and Portland Harbor. That project was not begun until much later in

1857, but congressional funding allowed for the construction of a far more impressive fortification.

Built of granite in a massive hexagonal shape, the modernized fort was described as imposing and impressive (Bradley 1981).

In each case, Congress heeded Maine’s call for visible long-range fortifications in the wake of

British aggression in the War of 1812 and the Aroostook War. Forts constructed on the coast of

Maine in those years served nation-building purposes by protecting crucial trade routes and by demonstrating to Mainers that their union would come to their aid if necessary. Many coastal fortifications, in particular, were constructed as broad, systemic projects including the “Third

System” plan to construct 30 fortifications between 1816 and 1867. Forts built under this initiative

12 were described as spectacular and impressive – visible for miles and providing a clear statement of the nation’s defensive readiness (Craig 1978, 68).

Elsewhere in the country, similar forts were constructed with the robust assistance of the national government, for a variety of nation-building purposes. For example, Congress appropriated funds for two major fortifications on the San Francisco Bay in the 1850s: Fort Point and Fort

Alcatraz. California was governed by military officers between 1846 and 1849 whose primary responsibilities included securing newly seized and settled lands and protecting settlers from Native

American tribes while thwarting filibustering missions into Mexico. Forts on the Atlantic coast and in the Great Lakes region were frequently constructed to assist in wartime operations and to defend

American borders from infringement by European powers. In the west, however, forts were tools of settlement and conquest as America sought an integrated empire stretching to the Pacific.

Following the gold rush, leaders shifted their resources towards the protection of the commercially vital Bay from foreign invasion and the guarantee of safe passage for ships importing goods. When

California achieved statehood in 1850, San Francisco Bay stood completely defenseless.

Constructing fortifications in prominent positions at the mouth and center of the bay allowed the federal government to clearly and visibly remind citizens of its presence in the state.

Like the military structures of Maine, investments in California security left an imposing physical presence. According to the National Park Service, Fort Point was described as “the pride of the Pacific,” “the Gibraltar of the West Coast,” and “one of the most perfect models of masonry in

America.” The fort on Alcatraz Island received an enormous sum of money in congressional appropriations between 1854 and 1857, and cannons and other fortifications were first installed on

Alcatraz Island in 1853 on orders of General John Wool who feared French threats. In 1855, threats of war with Spain led to increased appropriations for coastal fortifications across the country. While the fortifications on Alcatraz remained limited, citizens in San Francisco could certainly see the guns

13 and ringed armaments encircling the island, ready to defend the Bay should the fortification at Fort

Point fail. Ultimately, the investments in both Fort Point and Fort Alcatraz proved fruitful when southern sympathizers and western secessionists threatened to attack the city of San Francisco during the Civil War. Troops and guns stationed at each fort ensured that citizens were not lured into rebellion. When physical distance separates a state from the seat of national power, investments in military fortifications ward off threats of conquest and insurrection – the most dramatic means of political dissolution.

After passing two imposing forts and landing on the docks at Portland Harbor, merchants and visitors to Maine in the late 19th century would immediately recognize the distinct neoclassical architectural style of the new, federally funded Portland Custom House and the combined

Courthouse and Post Office building. Each received significant congressional appropriations between 1865 and 1874 to replace older, smaller buildings. Even a cursory look at the public buildings constructed in the 19th century makes clear that treasury department officials in charge of their design hoped to remind the American people of their government’s “extension to what are called matters of internal concern” (Hamilton, Federalist 27). Political and cultural leaders in this era strongly felt that architecture could serve as a tool of nation-building and moral education. Lowry

(1985) for example, quotes from an article appearing in the American Journal of Science and Arts in 1829 that suggested “objects of architectural beauty could provide a common bond of union among us; something to make us feel that we are members of one great community … and attach us more powerfully to our fellows and to the country.” Federal buildings, such as the Second Bank branch locations – so-called “money temples” – “established a federal presence and collectively approved architectural symbolism of stability and order, which could impart the aura of real stability” (Craig,

1978, 50).

Large, imposing structures commanded attention in the towns they occupied, raised the value of nearby commercial buildings, and provided visible reminders of the presence of the federal government. That was certainly the case in Portland. Moreover, local commissioners preferred neoclassical designs precisely because they served as “a means of attracting people and investment since they assumed that both would be drawn to towns that appeared civilized and prosperous”

(Bluestone, 1990, 133). Mainers, like citizens across the country, could recognize the tell-tale columns and porticos of federal buildings that literally and physically brought Washingtonian architecture, and a variety of accompanying benefits, to them.

Funding for prominent, elaborate customs houses and other public buildings was by no means exclusive to Maine. To this day, the Charleston Customs House in South Carolina stands as an imposing façade surrounded by much smaller buildings alongside the port. Like in Portland, a ship upon entering Charleston harbor for much of the 19th century would be faced first with the dual gates of Forts Sumter and Moultrie, and then by the impressive façade of the Customs House on the harbor front. That each of these buildings was federally funded and reflected the uniform architecture of federal construction is important for our understanding of the role they may have played in federal nation-building. Citizens of New Orleans could similarly boast of an enormous, neoclassical custom house. Built of granite with an expansive interior marble hall, that building’s original design plan called for exterior niches to be filled with statues of great American heroes. It is hard to imagine a more direct way of reminding residents of their connection to the nation.

The city of Portland, Maine provides a window into the thinking of 19th-century political leaders engaged in the process of allocating funding for internal improvements. From harbor and river improvements to forts and imposing federal buildings, these projects plausibly promised to bring a city on the literal edge of the nation into close contact with the country’s political and economic center. Internal improvements offered valuable infrastructural investments that connected

15 citizens, while integrating manufacturing and resource centers into a national economic system. By demonstrating a commitment to military defense, the federal government similarly signaled to the

American people that they derive key benefits from the constitutional union. Finally, as Hamilton suggested in the Federalist, a visible government always has an easier time converting disciples than an invisible one. The federal government deployed resources to conduct business in cities across the union in grand fashion. Internal improvements stimulated local economies by enlivening central squares and erecting ornate public buildings featuring Washington’s signature style. Such investments not only gave citizens a local source of national pride, they offered a crucial step towards economic and political integration.

Research Design

So far, we have presented the problem of conquering space as a particularly American challenge to nation-building and offered a few brief illustrations of internal improvement projects across the country. We turn now to an analysis of 19th-century data to evaluate if federal funding was more likely to flow into political communities disconnected from the union. To test this claim, we rely heavily upon the Gordon and Simpson (2018) measure of internal improvement appropriations.

Gordon and Simpson (2018) leverage a comprehensive record of over 8,000 local appropriations for public buildings, river and harbor improvements, forts, arsenals, armories, lighthouses, and a variety of other public works projects – a heroic geolocating and data cleaning effort made possible by an official Treasury Department report. Their measure codes federal spending specific to each congressional district in each congressional session between 1789 and

1882. The granularity of their data – district-level spending – is particularly useful, as congressional districts have roughly similar populations due to the constitutional guidelines for apportionment in the U.S. House. For this reason, we feel comfortable with our decision not to convert

16 appropriations into per-capita values. We adopt this measure with two minor modifications. For comparability across our time series, we rely upon consumer price index (CPI) estimates to adjust nominal appropriations dollars to 2018 values. We further recode this measure to indicate millions of U.S. dollars for ease of interpretability across our regression results. In short, we analyze the distribution of internal improvement appropriations, in millions of 2018-adjusted dollars, across each congressional district throughout much of the 19th century.2

In contrast to our outcome variable, identifying the most disconnected political communities proved to be quite a challenge. The conditions that necessitate greater attention to nation-building of this type almost certainly stem from a complex interaction of geographic, cultural, and political factors. Moreover, the difficulty in measuring something as abstract as “need for nation-building” is further compounded by the binding limitations and inevitable messiness of working with 19th- century observational data. Cognizant of these obvious limitations, we use each state’s physical distance from the national capital to construct two proxies for local nation-building demand.

Geographic distance from Washington, D.C. provides a useful foundation for these proxies in part because lawmakers explicitly invoked geographic considerations in their rhetorical appeals for internal improvement spending. Proponents of these federal projects routinely argued that distance

– mere physical separation – created a critical barrier to national unity. Disconnection is the central problem of nation-building in this vein. While it is certainly possible to be disconnected from the national community while residing at its center, problems of communication, transportation, and the development of an interdependent national economy are greatly exacerbated by physical distance.

2 For a more complete discussion of these data, please see the appendix materials provided by Gordon and Simpson (2018). We rely upon a formula provided by the Federal Reserve Bank of Minneapolis to adjust 19th-century dollars to 2018 values. Their consumer price index estimates incorporate five distinct data series from 1800 to the first quarter of 2018. Note that the U.S. dollar experienced considerable deflation between 1816 and 1882. Expenditures are adjusted for comparability and ease of interpretation. 17

Consequently, we argue that this proxy usefully captures the problem 19th-century American political leaders identified as central to their nation-building efforts.

Grounding our measurement strategy in geographic distance is also useful because it bundles together a variety of other, related attributes that suggest a need for national investment. Distant states during this period are more likely to be sparsely populated, to possess expansive territories lacking basic levels of infrastructure, and to border foreign nations and territories. Each of these attributes speaks to the broader explanatory factor we wish to identify – demand for integration.

Distance from the federal capital usually captures these correlated, difficult-to-measure features while continuing to take seriously – even literally – the rhetorical justification offered by lawmakers of the era.

Given these benefits, we record the total number of miles, by walking route, between each state’s capital city and the capitol building in Washington, D.C.3 We expect that congressional districts located within states further removed from the federal capital are likely to receive a disproportionate share of internal improvement investments; conversely, the inclusion of new states with a greater demand for nation-building will lead to a reduction in federal dollars appropriated for older, more established states. Table 1 indicates all states that are, at some point, among the top five most distant state in a given congressional session. Note that this distance measure is not merely a relabeled identifier for a particular geographic region, either; the western frontier states, for example, are unsurprisingly well-represented in the table below, but we also see a variety of states in the South and Northeast.

3 We estimate distance by Google Maps walking route, rather than by driving route or direct overland distance to better account for geographic barriers present in 19th-century transportation. While distances based on historical road maps would be ideal, walking routes come closest to approximating the relative ease of communication between states and the capitol. 18

Table 1. Distant States in the 19th-Century Miles from Sessions Among Top 5 State National Capitol Most Distant States Alabama 790 30 Arkansas 1,018 38 California 2,708 17 Colorado 1,634 14 Florida 796 2 Illinois 757 37 Indiana 567 5 Iowa 991 25 Kansas 1,123 26 Kentucky 526 22 Louisiana 1,118 75 Maine 592 6 Massachusetts 437 18 Minnesota 1,074 6 Mississippi 978 32 Missouri 923 18 Nebraska 1,158 18 Nevada 2,574 35 Ohio 395 2 Oregon 2,750 47 South Carolina 454 22 Tennessee 666 42 Texas 1,495 50 Vermont 492 17

We use this baseline distance calculation to construct two measures of nation-building demand that vary over time and place. First, we measure each state’s ranked proximity to the national capitol. Congressional districts located in more proximate states – i.e., those located near

Washington, D.C. – receive a high value for ranked proximity, while the most distant states are given lower numerical values. We expect that as a state’s relative proximity to D.C. increases – i.e., other states join the union in greater need of nation-building – congressional districts within that state will receive fewer internal improvement appropriations.

Next, we estimate the proportionate distance between state capitals and the national capital.

More specifically, proportionate distance takes the previously described distance measure (i.e., miles, in walking distance, between the state and national capital) and divides that value by the same value for the most distant state in a given congressional session. This measurement strategy thus codes the most distant state in each congressional session to have a proportionate distance value of 1, with other states’ distances measured in terms relative to the full scope of American political geography.

Unlike ranked proximity, this measure moves beyond an ordinal estimation of distance, while still allowing for variance over time and place. Moreover, this continuous measure allows us to incorporate the changing scale of the American republic and each state’s relative place within the growing nation.

We estimate the importance of physical distance on patterns in internal improvement appropriations using ordinary least-squares regression models that include both congressional session and district fixed effects.4 Congressional session fixed effects provide a useful means of accounting for unit-invariant shocks to all regions that might impact patterns in federal spending.

For example, congressional session (i.e., time) fixed effects allow us to account for shifts in baseline spending if we believe that Congress was uniformly less likely to spend money on certain projects in the legislative session following the outbreak of war.5 By contrast, district fixed effects provide an invaluable means of accounting for time-invariant characteristics that might influence internal improvement spending. For example, district fixed effects capture the presence of mountains,

4 We adopt Gordon and Simpson’s (2018) use of unique geographic geometries as district fixed effects, rather than rely upon potentially misleading measures based on proper names (e.g., VA-02 over time). Both sets of fixed effects are critical to controlling for difficult-to-measure confounding variables, but we continue to find statistically significant support for distance-based measures after excluding either time or unit fixed effects. 5 The important point here is that these are unit-invariant shocks. This is a helpful, but still imperfect, means of accounting for the effect of major political developments. For example, there is certainly a differential effect of southern secession on internal improvement spending across districts – a large swathe of districts simply left the Union. See the appendix for robustness checks on the Confederacy years and other late 19th-century developments. 20 climate, and other geographic features that consistently exist within a district’s boundaries but may be otherwise difficult to measure with precision in the 1800s.

Our fixed effects estimation strategy helps with many of the challenges posed by data analysis in this period, but we still need to account for confounding variables that vary across both district and congressional session. To address these factors, we include a battery of control variables unaccounted for by our fixed effects. Following Gordon and Simpson (2018), we include a pair of control variables that consider the political persuasion of each district’s representative. Districts represented by members of the majority party or closer to the chamber’s ideological median may be able to reap a disproportionate share of benefits for his district, and because these variables may also be correlated with political geography, we account for these potentially confounding influences. The committee system may also provide districts with an advantage in securing internal improvement spending. This may pose a problem to our analysis, particularly if older states are better represented on key committees. Consequently, we include indicators for districts represented on the operative funding committees (i.e., Ways and Means or Appropriations, depending on the year), as well as committees with policy jurisdiction over internal improvement projects – commerce, military affairs, and public buildings and grounds.

Finally, we manually code data from Holt (1977) to account for state government expenditures throughout this period.6 These records of total state expenditures – which are based upon state treasury reports – allow us to control for the strategic considerations presented to federal lawmakers. Federal lawmakers may have grappled with the utility of nation-building efforts in a political region that simultaneously receives – or does not receive – investments from sub-national

6 Please see Holt (1977) to evaluate the complete dataset and a full discussion of his methodology. As the appendix shows, our results are not dependent on the inclusion of this control variable. While comparable studies of internal improvements do not control for state spending, we feel that this is an important factor to take into consideration. Spending differs across states, varies over time, and appears to be positively correlated with federally funded investments in similar regions (p<0.0001). 21 governments. For example, Larson (2001) suggests that some states sought out further economic and political integration by funding their own transportation infrastructure projects. Even ardent opponents to federal internal improvements, such as the Jacksonians, conceded the need for practical infrastructure projects. They sought to guard against the consolidation of federal power and perceived programmatic threats to state autonomy. By including this variable, we account for intentional or incidental nation-building done by state governments (and not captured by fixed effects). Moreover, our decision to rely upon aggregate government expenditures allows us to account for state investments in crucial transportation and development projects. By accounting for state spending, we can better estimate the effect of nation-building demand (using distance as a proxy) on federal spending.7

Results

At a glance, congressional districts within distant states seem to receive more money than others. Figure 1 takes a cursory look at the unconditional relationship between physical distance and appropriations. The average appropriations received by congressional districts in the most distant state is $631,212 (2018-adjusted), while the average among all other states is $240,462. Put differently, districts in the most distant state typically receive between two and three times as much funding, and this difference ($390,749) is statistically significant (p<0.0001). These results, however, ignore the relative power of a district’s representative, the physical geography of a region, and a variety of important factors that might complicate this relationship. Moreover, we feel that a narrow

7 The effective sample of our analysis is 1820-1882. While the source internal improvements data begins several decades earlier, we analyze this truncated time series for several reasons. First, we lack CPI data before 1800 to adjust our dependent variable. Second, the state expenditure data, coded from Holt (1977), only dates back to 1820. Finally, we follow Gordon and Simpson (2018), who drop all data before 1823 from their analysis “owing to the paucity of local appropriations” in this period (570). 22 focus on the single most distant state in a given year fails to fully account for the broader concept we wish to measure: demand for nation-building efforts.

Figure 1. Congressional Districts in Distant States Consistently Received More Funding

Note: Local polynomial smooth plots for congressional districts in the most distant state in each year (solid blue) of the dataset, compared to those in all other states (dashed black), with 95% confidence intervals.

Table 2 presents the results of our primary analyses. In each column, we include a more continuous measure of nation-building, the previously described battery of control variables, and a pair of fixed effects, and in both models, we find some support for our expectations. Congressional districts appear to receive more internal improvements funding when they are located in states that are, relative to others, more distant from D.C. Put differently, congressional districts receive fewer

23 federal dollars when other states are admitted to the union in regions of North America even further removed from, and thereby disconnected from, the seat of national government power.

Table 2. Distance is Correlated with Federal Investments DV: Appropriations (Millions of 2018-Adjusted USD) Ranked Proximity -0.05*** (0.02) Proportionate Distance 1.36* (0.70) Majority Party 0.004 -0.001 (0.03) (0.03) Ideological Distance to Median -0.14 -0.15* (0.09) (0.09) Funding Committee 0.03 0.03 (0.05) (0.05) Commerce 0.17* 0.17* (0.09) (0.09) Military Affairs -0.02 -0.01 (0.05) (0.05) Public Buildings and Grounds 0.31* 0.31* (0.17) (0.17) State Government Expenditures 0.003 0.003 (0.003) (0.003) District Fixed Effects Yes Yes Session Fixed Effects Yes Yes Observations 20,348 20,348 R-squared 0.36 0.36 OLS coefficients with robust standard errors in parentheses. (*** p<0.01, ** p<0.05, * p<0.1, two-tailed tests)

More specifically, the results in the first column indicate that reduced demand for integration and national connection, as measured by ranked-proximity, corresponds with a significant decrease in federal appropriations for internal improvement funding (p=0.001). A one standard deviation increase in ranked proximity corresponds with a $429,377 decrease in internal improvement funding, suggesting that districts benefit disproportionately when they exist within relatively remote states.

This is a substantial increase, given that the average value for our dependent variable is $244,204 and many districts receive no funding at all in a given year. In the second column, we continue to find support for our hypothesis, but we have slightly less confidence in our ability to reject the null

24 hypothesis of no relationship between distance and appropriations dollars. In this model, one standard deviation increase in proportionate distance corresponds with a $258,871 increase in internal improvement funding, but our estimate moves just beyond the traditional – if arbitrary – threshold for statistical significance (p=0.054). While the magnitude and significance of the results vary slightly by measurement strategy, both models suggest that the need for a specific type of nation-building may have been an important predictor for the receipt of government resources.

Taken together, these findings support the rhetoric advanced by proponents of internal improvement projects. We perform a variety of robustness checks in the appendix, and the interpretation of our results remains virtually the same. We cannot, however, make any definitive claims as to whether these programs actually “conquered space” in the ways articulated by the great orators of the 1800s; we have evaluated the distribution of money, but we are unable to estimate the impact of those dollars in the present context. Nevertheless, the results do suggest a pattern of internal improvement allocations consistent with the logic of federal nation-building – even after accounting for traditional measures of congressional power, time-invariant features of congressional districts (e.g., physical geography), temporal shocks, and a time- and unit-varying measure of state government expenditures not included in most analyses of internal improvement appropriations.

Still, we caution against taking these results as definitive proof of a causal relationship. As previously mentioned, our analysis requires a series of measurement assumptions and econometric techniques that rely heavily upon the accuracy of 19th-century data and best-available proxies. While we argue that distance is a useful proxy for nation-building demand, our statistical analyses necessarily obscure the varied mechanisms by which internal improvements may have served as a means to bind the nation together.

Discussion

We began this paper with the uncontroversial claim that America, at its founding, faced significant challenges. America seemed to occupy an ungovernably large domain, and the nation was determined to compound this problem with aggressive territorial expansion. Certainly, history suggested, such a vast region could never survive as a republic. At the same time, the Constitution’s mixed regime – partly national and partly federal – limited the nation-building tools available to political leaders. While a unitary regime might have conquered space through the imposition of uniform law, and a federal regime might leave distant states to their parochial attitudes, a mixed regime could do neither. America’s early political leaders needed a way to inculcate the nationalism and shared interests that would allow a large republican regime to survive without directly dictating terms to independent states.

The size of the republic thus posed a significant threat to the stability of the new American government, but that threat could be mitigated, many leaders claimed, through active projects of nation-building. Drawing the peripheries of the union into closer connection with the federal government, would better protect America’s borders from external threats, would connect citizens and their material interests in ways that overcame economic competition, and would increase the visibility of congressional action, reinforcing the rule of law and America’s commitment to a partly national, and not merely federal regime. Minicucci (2001) frames this perspective as “interest-based nation-building” and argues that “space was the enemy of the new American republic” and that

“true nationalism … is firmly rooted in the material process of economic integration through which individuals perceive far-flung national interests as their own” (248). The goal for early American leaders was to “help foster a shared sense of national interest that could supplant the shared material interests that occurred more naturally at the regional or sectional level” (ibid., 257). By exploiting key mechanisms of nation-building – e.g., transportation, security, and visible commercial investment –

26 projects of internal improvement presented a potential solution to the problem of conquering space in a mixed regime.

Despite these advantages, the high stakes of nation-building proposals invariably stoke political conflict, and throughout the 19th century, American legislators struggled over the benefits of much-needed infrastructure investments and the costs of federal consolidation. To advocates, internal improvements offered a sweeping promise of national unity. Federally funded enterprises, from public buildings to arsenals, were believed to overcome “the problem of space for a republic,” as political cohesion was expected to follow from economic integration (Minicucci 2001, 248).

Opponents, however, argued that these expenditures risked corruption and threatened the balance of powers in a delicate federal system. The rhetoric surrounding internal improvements was bold, ambitious, and constitutionally controversial, as the nation’s economic development policy would set the foreground for pitched sectarian conflict and partisan politics in postbellum America (Bensel

2000).

Nevertheless, much of the research on internal improvements has primarily focused on the presence, or absence, of distributive legislative behavior. In contrast to this research, we present a simple test of the motives put forth by internal improvement advocates. Using a dataset of 19th- century district-level appropriations, we find that patterns in federal appropriations are consistent with a nation-building strategy. Congressional districts at the fringes of the republic appear to have received disproportionate support from the federal government, even after accounting for important attributes of physical geography, legislative influence, and sub-national spending patterns.

These findings contribute to a growing body of research in American political development and economy. Moreover, this work suggests that both American and comparative federalism studies should move beyond a purely distributive focus and consider large, public-interest purposes. More work is needed, however, to bridge our legislature-centered perspective with new work that

27 incorporates the calculations of executive branch officials (Rogowski 2016a, 2016b). Moreover, additional archival research would allow for a more nuanced evaluation of district-by-district variance in the demand for internal improvements. More specifically, we have not fully grappled with territorial politics, nor have we engaged with sovereignty issues around slavery and anti- consolidation resistance (Mulcare 2008). Future researchers should also attempt to evaluate asymmetric, centralizing shifts in political authority among those states that disproportionately benefited from internal improvement projects (Kincaid 2019). Finally, few political scientists have systematically analyzed the political consequences of internal improvements (but see Rogowski and

Gibson 2016, for an example of this approach applied to post offices). While we find suggestive evidence that projects were funded according to a nation-building perspective, the effectiveness of these projects remains an open question. It is possible, in other words, that a “cement of interest” was applied, but did not hold, at a critical time in the American republic.

References

Adler, William D. 2012. “State Capacity and Bureaucratic Autonomy in the Early United States: The Case of the Army Corps of Topographical Engineers.” Studies in American Political Development 26(2): 107-124. Bensel, Richard Franklin. 2000. The Political Economy of American Industrialization, 1877-1900. New York: Cambridge University Press. Blanchard, Newton C. 1894. “River and Harbor Improvement” The North American Review 158(448): 343-352. Bluestone, Daniel. 1990. “Civic and Aesthetic Reserve: Ammi Burnham Young's 1850s Federal Customhouse Designs” Winterthur Portfolio, 25(2/3), 131-156. Borneman, Walter R. 2004. 1812: The War that Forged a Nation. New York: Harper Collins. Bradley, Robert L. 1981. “The Forts of Maine, 1607-1945: An Archeological and Historical Survey.” Maine Collection. 6. Calhoun, John C. 1888. “Speech on the Bill to Set Aside the Bank Dividends and Bonus as a Permanent Fund for the Construction of Roads and Canals.” In The Works of John C. Calhoun. New York: D. Appleton. v.2. Craig, Lois A. 1978. The Federal Presence: Architecture, Politics, and Symbols in United States Government Building. Cambridge, Mass: MIT Press. Ellis, Richard E. 1987. The Union At Risk: Jacksonian Democracy, States’ Rights and the Nullification Crisis. New York: Oxford University Press. Finocchiaro, Charles J. 2015. “The Public Buildings Boom: Distributive and Partisan Politics in the Modernizing Congress” working paper: https://www.vanderbilt.edu/csdi/events/Finocchiaro.pdf Finocchiaro, Charles J. and Jeffery A. Jenkins. 2016. “Distributive Politics, the Electoral Connection, and the Antebellum U.S. Congress: The Case of Military Service Pensions.” Journal of Theoretical Politics 28(2): 192-224. Goodrich, Carter. 1960. Government Promotion of American Canals and Railroads, 1800-1890. New York: Columbia University Press. Gordon, Sanford, and Hannah Simpson. 2018. “The Birth of Pork: Local Appropriations in America’s First Century.” American Political Science Review 112(3): 564-579. Holt, Charles F. 1977. The Role of State Government in the Nineteenth Century American Economy, 1820- 1902. New York: Arno Press. Jenkins, Jeffery A. and Sean Gailmard. 2018. “Distributive Politics and Congressional Voting: Public Lands Reform in the Jacksonian Era” Public Choice. 174(3-4): 259-275. Kincaid, John. 2019. “Dynamic De/Centralization in the United States, 1790-2010.” Publius: The Journal of Federalism, 49(1): 166-193. Larson, John Lauritz. 2001. Internal Improvement: National Public Works and the Promise of Popular Government In the Early United States. Chapel Hill: University of North Carolina Press. Longnecker, Julia Ward. 1985. “A Road Divided: From Memphis to Little Rock through the Great Mississippi Swamp.” Arkansas Historical Quarterly 44(3): 203-219. Lowry, Bates. 1985. Building a National Image: Architectural Drawings for the American Democracy: 1789- 1912. Washington, D.C.: National Building Museum.

Madison, James. 1815. “Seventh Annual Message.” Online by Gerhard Peters and John T. Woolley. The American Presidency Project. https://www.presidency.ucsb.edu/node/204622. Madison, James, Alexander Hamilton, and John Jay. 1961. The Federalist Papers, ed. Clinton Rossiter. New York: Signet. Malone, Laurence. 1998. Opening the West: Federal Internal Improvements Before 1860. Westport: Greenwood Press. Minicucci, Stephen. 2004. “Internal Improvements and the Union, 1790-1860.” Studies in American Political Development 18(2): 160-185. Minicucci, Stephen. 2001. “The ‘Cement of Interest’: Interest-Based Models of Nation-Building in the Early Republic.” Social Science History 25(2): 247-274. Mulcare, Daniel M. 2008. “Restricted Authority: Slavery Politics, Internal Improvements, and the Limitation of National Administrative Capacity.” Political Research Quarterly 61(4): 671-685. Pears, Emily. 2017. “Chords of Affection: A Theory of National Political Attachments in the American Founding.” American Political Thought 6(1): 1-29. Remini, Robert V. 1991. Henry Clay: Statesman for the Union. New York: W.W. Norton & Company. Rogowski, Jon C. 2016a. “Mobilization and the Mail: Voter Turnout and the Expansion of the U.S. Postal Service.” working paper: http://scholar.harvard.edu/files/rogowski/files/mobilization_mail.pdf Rogowski, Jon C. 2016b. “Presidential Influence in an Era of Congressional Dominance.” American Political Science Review 110(2): 325-341 Rogowski, Jon C. and Chris Gibson. 2016. “Returns to Sender? Ballot Reform, Partisan Institutions, and the Electoral Consequences of American State-Building” working paper: http://scholar.harvard.edu/files/rogowski/files/australian_ballot.pdf Sheehan, Colleen A. 2009. James Madison and the Spirit of Republican Self-Government. New York: Cambridge University Press. Wilson, Rick K. 1986. “An Empirical Test of Preferences for The Political Pork Barrel: District Level Appropriations For River And Harbor Legislation, 1889-1913.” American Journal of Political Science 30(4): 729-754.

Online Appendix

Table A1. Ranked Proximity Results DV: Appropriations (Millions of 2018-Adjusted USD) Ranked Proximity -0.05*** -0.05*** -0.05*** -0.05*** (0.01) (0.01) (0.01) (0.02) Majority Party 0.01 0.01 0.004 (0.02) (0.02) (0.03) Ideological Distance to Median -0.13* -0.12 -0.14 (0.08) (0.08) (0.09) Funding Committee 0.02 0.03 (0.04) (0.05) Commerce 0.15* 0.17* (0.08) (0.09) Military Affairs -0.01 -0.02 (0.05) (0.05) Public Buildings and Grounds 0.31* 0.31* (0.16) (0.17) State Government Expenditures 0.003 (0.003) District Fixed Effects Yes Yes Yes Yes Session Fixed Effects Yes Yes Yes Yes Observations 22,974 22,935 22,935 20,348 R-squared 0.36 0.36 0.36 0.36 OLS coefficients with robust standard errors in parentheses. (*** p<0.01, ** p<0.05, * p<0.1, two-tailed tests)

Note: Table A1 illustrates that our full model (see far right column) is not dependent on the inclusion of state expenditures, committee assignment, or ideological control variables.

Table A2. Proportionate Distance Results DV: Appropriations (Millions of 2018-Adjusted USD) Proportionate Distance 1.01* 1.00* 0.99* 1.36* (0.52) (0.52) (0.51) (0.70) Majority Party 0.01 0.01 -0.001 (0.02) (0.02) (0.03) Ideological Distance to Median -0.14* -0.12 -0.15* (0.08) (0.08) (0.09) Funding Committee 0.02 0.03 (0.04) (0.05) Commerce 0.15* 0.17* (0.08) (0.09) Military Affairs -0.01 -0.01 (0.05) (0.05) Public Buildings and Grounds 0.31* 0.31* (0.16) (0.17) State Government Expenditures 0.003 (0.003) District Fixed Effects Yes Yes Yes Yes Session Fixed Effects Yes Yes Yes Yes Observations 22,974 22,935 22,935 20,348 R-squared 0.36 0.36 0.36 0.36 OLS coefficients with robust standard errors in parentheses. (*** p<0.01, ** p<0.05, * p<0.1, two-tailed tests)

Note: Table A2 illustrates that our full model (see far right column) is not dependent on the inclusion of state expenditures, committee assignment, or ideological control variables.

Table A3. Results Hold after Dropping Military Appropriations DV: Non-Military Appropriations (Millions of 2018-Adjusted USD) Ranked Proximity -0.04*** (0.01) Proportionate Distance 1.10* (0.61) Majority Party 0.01 0.003 (0.02) (0.02) Ideological Distance to Median -0.09 -0.10 (0.08) (0.08) Funding Committee 0.001 -0.001 (0.04) (0.04) Commerce 0.07 0.07 (0.08) (0.08) Military Affairs -0.04 -0.04 (0.04) (0.04) Public Buildings and Grounds 0.37** 0.37** (0.17) (0.17) State Government Expenditures 0.003 0.003 (0.003) (0.003) District Fixed Effects Yes Yes Session Fixed Effects Yes Yes Observations 20,348 20,348 R-squared 0.34 0.34 OLS coefficients with robust standard errors in parentheses. (*** p<0.01, ** p<0.05, * p<0.1, two-tailed tests)

Note: Table A3 illustrates that our primary results are not exclusively driven by military projects of internal improvement.

Table A4. Results Hold after Excluding Years After 1873 DV: Appropriations (Millions of 2018-Adjusted USD) Ranked Proximity -0.05*** (0.01) Proportionate Distance 1.31* (0.71) Majority Party 0.01 0.01 (0.02) (0.02) Ideological Distance to Median -0.06 -0.08 (0.07) (0.07) Funding Committee 0.03 0.03 (0.05) (0.05) Commerce 0.19* 0.19* (0.11) (0.10) Military Affairs 0.03 0.03 (0.05) (0.05) Public Buildings and Grounds 0.18 0.17 (0.23) (0.23) State Government Expenditures 0.003 0.003 (0.004) (0.004) District Fixed Effects Yes Yes Session Fixed Effects Yes Yes Observations 15,549 15,549 R-squared 0.32 0.32 OLS coefficients with robust standard errors in parentheses. (*** p<0.01, ** p<0.05, * p<0.1, two-tailed tests)

Table A4 illustrates that our primary results are not exclusively driven by major developments in the late 19th century that may have differentially affected congressional districts (e.g., the panic of 1873, the rising importance of railroads). Results are similar, and in some cases improved, if we exclude all years after 1865, which indicated a boon in railroad construction.

Table A5. Results Hold after Excluding 1861-1870 DV: Appropriations (Millions of 2018-Adjusted USD) Ranked Proximity -0.06** (0.03) Proportionate Distance 1.36* (0.74) Majority Party 0.04* 0.04 (0.02) (0.02) Ideological Distance to Median 0.03 0.03 (0.08) (0.08) Funding Committee 0.03 0.03 (0.06) (0.06) Commerce 0.23** 0.23** (0.11) (0.11) Military Affairs -0.04 -0.04 (0.05) (0.05) Public Buildings and Grounds 0.12 0.12 (0.13) (0.13) State Government Expenditures 0.002 0.002 (0.003) (0.003) District Fixed Effects Yes Yes Session Fixed Effects Yes Yes Observations 16,748 16,748 R-squared 0.45 0.45 OLS coefficients with robust standard errors in parentheses. (*** p<0.01, ** p<0.05, * p<0.1, two-tailed tests)

Note: Table A5 illustrates that our primary results are unlikely to be confounded by the Confederate years, which may have differentially affected congressional districts. While this period can be measured in various ways, we begin with southern secession in 1861 and end with the full readmission of southern states to Congress in 1870.