The Purpose and Powers of the Senate, Part XXII: “Continuity of Service”

    As of April 8^th, 1913, the 17^th Amendment was officially ratified. The procedure by which United States Senators had been chose since the Constitution was adopted in 1788 was thereafter replaced with a system of popular elections, thus effectively removing the state assemblies from any substantive role in the federal power structure. The first competitive election to take place under this new system occurred in Maryland in November of that same year, the result of the sudden death of Democrat Isidor Rayner (1850-1912). The Democrats held the seat. The first nationwide election to take place under its auspices occurred the following year, in November of 1914. The results, in spite of the nature of the modification that had been wrought, were also rather unspectacular. Of the thirty-two Senate seats that were contested by way of popular elections, only three of them changed hands as a result of a loss at the polls. The majority of incumbents kept the seats they had been previously granted; of those races which were open – namely as the result of retirement – the majority resulted in no change in party status. Elections for the two remaining classes of Senators were somewhat more dynamic. In 1916, a full twelve seats out of thirty-five changed hands. But then, in 1918, that number went down to eight out of thirty-eight. Based on these short-term results alone, very little seemed to have changed. The inhabitants of the various states were evidently inclined to make much the same choice for Senator as their legislatures had before. The chamber was not thrown into utter turmoil. The parties did not experience sudden reversals of fortune. It was almost as if nothing had happened at all.

    This was not the case, of course. Something of fundamental significance to the very nature of republican government in the United States had in fact just taken place. The reason that its impact was not immediately observable was that one requires access to the right data in sufficient amounts to be able to properly observe just what changed and how. And while this kind of data may not have existed in 1914, or 1916, or 1918, it most definitely does exist now. All credit for compiling this information is due to those scholars of the Senate whose efforts have succeeded in producing an exceptionally thorough portrait of that chamber’s structural evolution following the passage of the 17^th Amendment. In particular, recognition must be granted – for his academic work if not for his other professional “accomplishments” – to United States Circuit Court Judge Jay Bybee (1953-present) who published among the first comprehensive studies of the effects of the 17^th Amendment in an issue of the Northwestern University Law Review (Vol. 92, No. 2) in 1997. Bybee’s work provides the basis for the bulk of conversation that follows.

    So what, then, was the “actual result” of the ratification of the 17^th Amendment? In what way, if any, did this modification of the Constitution change the Senate of the 19^th century into the Senate of the 20^th century? The answer, unsurprisingly, is more than a little complex. To begin with, when one gathers the relevant data on the political composition of the various state legislatures between the late 1860s and the late 1990s and compares it to the record of political representation in the Senate on a state-by-state basis, two significant trends immediately present themselves. In the first case, between 1869 and 1914 – when the first nationwide election was held under the terms of the 17^th Amendment – the political affiliation of a given state’s Senators almost always reflected the political composition of that same state’s legislative assembly. If a state assembly, at a given moment, was controlled by the Democrats, for example, then the Senator which that assembly chose would almost always be a Democrat as well. Granted, this is not much of a conclusion on its own. It would seem only natural that the partisans who found themselves in control of a state legislature would seek to extend their power into the federal realm as well. But consider, in addition, the following. After the ratification of the 17^th Amendment in 1913, control over the various state legislatures no longer had any causal impact upon the partisan makeup of the Senate. Pre-1914, the party that controlled the majority of the state assemblies was essentially guaranteed to enjoy a proportionate majority in the Senate. After 1914, this was no longer the case. The Senate and the state assemblies thus became formally unlinked, allowing the two to diverge and evolve in ways that their continued association would almost certainly not have allowed.

    In the most baseline, practical sense, this divergence has given rise to drastically different outcomes in terms of federal legislative elections. Take the aforementioned election of 1914 as a case in point. While, on the surface, the ratification of the 17^th Amendment seemed to exert little change, if any, on the final outcome – the Democrats gained three seats and remained in control of the chamber – the likely outcome in its absence rather belies this superficial conclusion. Recalling that, prior to 1914, the party affiliation of any given Senator almost always reflected the party affiliation of the controlling majority in the relevant legislature, a 1914 Senate election in which the state assemblies continued to appoint the membership of the upper chamber would almost certainly have given rise to a far more tenuous balance of power. Under the terms of the 17^th Amendment, the election of 1914 gave rise to a Democratic majority in the Senate by a margin of 56-42. In the absence of a 17^th Amendment, however, the result would more than likely have been an even split of 48-48. In actual fact, the Democratic-controlled Senate of the 64^th United States Congress undertook the passage of a number of very consequential pieces of federal legislation addressing a wide variety of policy areas. These included, but were not limited to, the Tillman Act (which, for the first time, sought to regulate campaign finance), the Kent-Smoot Act (which established the National Park Service), the Jones Act (which provided for an elected legislature in the occupied Philippines), and the Army Appropriations Act (which laid the groundwork for the modern military-industrial complex). While it would be impossible to say for certain that in the tense climate of a 48-48 Senate none of these laws would have been passed, the legislative priorities of both the Democrats and the Republicans were sure to have been quite different when compared to the actual events of the 64^th Congress. It would accordingly seem more than fair to conclude that by unhooking the composition of the Senate from that of the various state assemblies, the 17^th Amendment began to change the course of American history as early as 1914. Laws which might not have passed were approved; law which might have been passed were not.

    And this was far from the only occasion in which the actual political outcome significantly diverged from what might have been the case in the event that the 17^th Amendment was never enacted in 1913. In 1932, for example, whereas in reality the Democrats took possession of the majority in the Senate, the absence of the 17^th Amendment would have likely left it in the hands of the Republicans until 1934 at the least. Doubtless this would have offered a significant challenge to President Roosevelt’s efforts to secure passage of a host of early New Deal legislation, with the possible outcome of a significantly worsened Great Depression. Just so, without the 17^th Amendment, there almost certainly wouldn’t have been a Democratic Senate majority in place in the late 1940s and early 1950s, thus likely complicating the ratification of the original NATO charter in 1949, the passage of the Immigration and Nationality Act of 1952, and the approval by Congress of the 22^nd Amendment in 1951. Even more recent events would almost certainly have turned out differently. The Republican Party would still have taken back majority control of the House in 1994 for the first time in forty-two years, but the continued appointment of Senators by the state assembles would have allowed the Democrats to actually make gains in the Senate, increasing their majority from fifty-seven seats to seventy. In the event of such an alternative outcome, one wonders whether the government shutdowns of November, 1995 and December, 1995 would have taken place at all given the ability of a Senate controlled by the Democrats to hold a Republican-controlled House in check.

    As far as the upper chamber’s character is concerned, it might be fair to say that the average Senator has become both more independent and less independent than was previously the case. On the one hand, before the ratification of the 17^th Amendment, Senators were subject to being directly instructed by the state assemblies that appointed them. This was not a formal power which the assemblies possessed, it bears noting, but rather a logical outgrowth of the relationship between the two. If a Senator declined to follow a set of written instructions provided by the members of the legislature that had chosen them, these same members would seem to have little cause re-appoint the Senator in question when their tenure in office expired. The result, in practice, was that Senators generally did as they were told. Following the adoption of the 17^th Amendment, this was no longer to be the case. Since then, Senators have only been beholden to the whims of their home state’s population. And even then, only in the most vague and ill-defined sense. They can and do take polls of their constituents in an attempt to ascertain their feelings on a given issue or policy, but for the most part they simply vote as they feel it is in their interest to do so. Rather than receiving instructions, therefore, Senators since the adoption of the 17^th Amendment had arguably taken to giving explanations. While on the Senate floor, they vote; while on the campaign trail, they justify. Granting that their explanations do not always pass muster, the general dynamic would still seem to grant them a markedly freer hand. With no explicit instructions to risk their career by disobeying, Senators have been given room to negotiate the terms of their continued tenure in office with a larger and less cynical constituency than the membership of a state assembly. Career politicians, after all, are less likely to be vulnerable to campaign rhetoric and dramatic promises than the average American voter. A state assemblyman must be reasoned with; an American citizen can be convinced.  

    On the other hand, Senators have arguably become even more beholden to the general electorate than they had ever previously been obligated to the membership of the state assemblies. Consider, to that end, the nature of the change in question. Before the adoption of the 17^th Amendment, any given Senator was only ever directly accountable to a majority of the members of their home state’s legislative assembly. Depending on the state, of course, this could range from a fairly small number of people to a very large one. In either case, however – and in every case in between – the rules of the pre-1913 Senate rendered the number of people who actually elected a given Senator into something that was both specific and cognizable. So long as a Senator had the trust and support of this specific group of people – by way of promises, good behavior, or perhaps an exchange of favors – their continuation in office was more or less assured. Assured, that is, so long as the legislators in question maintained their majority. Senators played no small part in this as well, it bears noting. Many a state legislative election, prior to the ratification of the 17^th Amendment, pivoted upon the question of whom the candidates proposed to appoint to the Senate. That being said, the intermediate role played by the state assemblies did still significantly color the resulting dynamic. Assemblymen had no reason to simply appoint whoever was most popular with their constituents, though popularity no doubt factored into their thinking. Rather, they were also obliged to consider who among the possible candidates for Senate would be the easiest to instruct, the most loyal, and the most likely to represent their state’s interests in good faith. In consequence, while Senators prior to 1913 most certainly did undertake to publicly campaign for their seats – a phenomenon perhaps most famously exemplified by the Lincoln-Douglas Debates of 1858 – their most important relationship was with the legislators who directly appointed them. 

    Since the ratification of the 17^th Amendment, however, Senators have been forced to become popular electioneers. To some extent, as aforementioned, this has given them a freer hand in terms of pleading their case and successfully swaying their constituents. The average citizens, it bears repeating – without judgement or disdain – does tend to be more susceptible to political rhetoric than the average elected official. But it has also resulted in Senators paying far greater attention to the more facile aspects of public service. Convincing a group of legislators to send them to the Senate no doubt demanded a great deal of a candidate in the way of arguments and promises, favors and pledges of gratitude. But it was not an effort that necessarily required their constant and unwavering attention. Being beholden to the general population, by comparison, would seem to turn every public gesture, every vote, and every address into potential material for later campaigning. The members of a state assembly, after all, didn’t need to pay heed to the news coverage to know what it was their Senator stood for or how they intended to vote. In need of information, they could simply ask the individual to make a report. And desirous of a specific outcome, they could simply deliver to them formal instructions. But because the general population doesn’t have access to, or control over, their chosen Senators in this way, these same Senators must make a point of essentially broadcasting a desirable image essentially any time they are in the public eye. In essence, they must chase votes, rather than being able to simply gather them. They must cast as wide a net as possible, rather than seek out the relatively modest degree of support which they need for their careers to survive. And this, in its own strange way, has made Senators significantly less autonomous than was previously the case.

    As to the questions of tenure and rotation, the 17^th Amendment has also given rise to some rather unexpected developments. Prior to 1913 – generally speaking – the length of service of most Senators tended to be fairly short. This isn’t to say that most Senators failed in their reelection bids or that no Senator ever secured a string of several consecutive terms in office. Rather, it is simply to point out the basic fact that Senate tenures, on average, have become far longer since the adoption of the 17^th Amendment than they ever were before. Consider, by way of an example of how things used to be, the careers of the three men who have long been heralded as the greatest United States Senators of 19^th century. Henry Clay (1777-1852), who racked up no less than four separate stints in the upper house over the course of his lengthy career in public service, was only ever elected to a consecutive term once, in 1836. Prior to that, he served for less than a year between 1806 and 1807 and for slightly more than a year between 1810 and 1811. At the end of his life, in 1849, he was elected once again, only to then die in office in 1852. All told, this would seem to amount to two full terms and a little more than half of a third. John C. Calhoun (1782-1850) was even less prolific than Clay, managing the better part of a single term between 1832 and 1843 and then another partial term between 1845 and his own death in 1850. Daniel Webster (1782-1852), by comparison with these two, had a much more stable tenure in office. Between 1827 and 1841, he managed to string together three consecutive elections. This was followed by a fourth term in 1845, after which point he retired to become Secretary of State. Clay and Calhoun, to be fair, also cut their Senate careers short at times for reasons both personal and professional. Calhoun did so in 1843, for example, in exchange for the leadership of the State Department. And Clay’s second and forth terms both came to an end voluntarily. But it is nevertheless exceptionally telling how short the tenures of these three “great men” actually were. The did not park themselves in a Senate seat and serve for decades on end, but rather came and went, were reelected and resigned, in accordance with their own personal desires and those of the legislators to whom they were beholden.

    The reason that this kind of churn was once fairly common most assuredly has to do with various ambitions of those involved. Senators have, since 1913, resigned their seats voluntarily in order to accept a federal appointment, or done so out of a desire to retire from public life on their own terms rather than face eventual defeat. But rarely, once they’ve made such a choice, do they then return to office years later. Political careers have become longer, on average, but they have also become less dynamic. The 17^th Amendment has contributed to this development in the way that it altered the relationship between the different tiers of popular government. Prior to its ratification, there arguably existed a kind of pipeline between the state governments and the federal government by which ambitious public servants attempted to advance their careers and accrue favors. Through election to a state assembly, one gained influence over Senate appointments; through voting on Senate appointments, one accumulated connections and traded favors; by trading favors, one built up a case for being elected to the Senate oneself; and by being elected to the Senate, one gained access to federal appointments and federal patronage by which to secure eventual reelection. Constant upward pressure essentially created a kind of conveyor belt, with legislators seeking to advance themselves while checking the careers of those they supported and Senators seeking access to federal appointments so as to reward those who had elevated them in turn. In this way, while the members of a given state assembly might have preferred not to continually reelect the same Senator indefinitely – specifically because many of them naturally coveted the seat in question – a former Senator who had been elevated to a position in the Cabinet might nonetheless achieve their own reelection by way of the judicious application of patronage.

    By making United States Senators subject to popular election, the ratification of the 17^th Amendment has fundamentally altered this dynamic. Ambition has still come to define the trajectory of many a career in public service, to be sure, but these trajectories are a fair bit more linear than they used to be. And while the desire for professional advancement continues to propel individuals from one office to another in search of greater wealth and prestige, the shift from legislative appointment to popular election in the Senate has eliminated what was once a major junction point between the state and federal stages of the American political career track. Previously, election to a seat in any of the state assemblies more or less guaranteed a person access to some portion of the federal power structure. By casting their vote for Senator, state legislators built up relationships with their federal counterparts and accrued favors with their fellow lawmakers, thus paving the way for potential patronage appointments – if the Senator they elected happened to gain appointment to the Cabinet – and helping to establish the necessary powerbase to accomplish their own election to the Senate in turn. Following the ratification of the 17^th Amendment, however, the jump from state service to federal service has become significantly more difficult. Rather than being able to use the appointment power as a kind of springboard to federal office, individuals have instead taken to building up a public profile in state government so as to gain the endorsement of a national party apparatus for a run for federal office. Individuals do still run for Senate without any prior experience, to be sure, particularly if their public profile is exceptionally impressive. But it remains most common for aspiring public servants to first spend a few terms in either the House of Representatives or the executive branch of their home state before then becoming a candidate for the upper chamber of Congress.

    The reason for all this is that a given Senator’s public profile has become by far the most valuable currency by which they are able to pay off their continuation in office. Their most important relationship is now with the voters rather than a handful of legislators, creating a much greater incentive to be seen and be heard than was ever the case before 1913. As far as their prospects for reelection are concerned, a Senator is essentially the culmination of their votes, their speeches, their public statements, and their presence in the media. The old priorities have been eliminated – i.e., the need to maintain good relationships with the members of the relevant state assembly – and so has the upward pressure that once encouraged Senators to serve briefly and then proceed to repay their debts. Now, so long as the voting public is sufficiently satisfied, there are few reasons for a member of the Senate to seek even higher office. Some may pine for the presidency, it is true, as a kind of capstone to their political career, but these are relatively few. Most modern Senators, having attained the public office in the whole of the American republic with the longest tenure between elections, settle in, pay heed to their campaign war chests, and begin planning for their eventual retirement. Their race to the top of the power structure having essentially been won, they embrace their purported status as the grandees of their respective parties.

Interestingly enough, this is not at all what some of the loudest critics of the 17^th Amendment predicted would occur when the topic of Senate election reform was still a matter of public debate. In 1902, for example, New York Senator Chauncey Depew (1834-1928) objected to any such reform on the grounds that Senators should not have had the, “Tenure of their place here” questioned every six years by people who lacked sufficient knowledge of the necessary qualifications. To his thinking, it seemed, allowing such popular review of the performance of sitting Senators was bound to result in an unhealthy degree of turnover at the expense of efficient and effective policymaking. Just so, writing in 1906, political scientist George H. Haynes (1866-1947) opined in his Election of Senators that a shift to popular election would produce even shorter tenures of service. “The choice of senators by state legislatures” he said,

Has tended to produce a continuity of service, and hence an efficiency based upon long experience in legislature work, highly exceptional in popular governments [...] If the effects of popular elections be judged by results produced in the election of governors and representatives in Congress, it is clear that the trading of localities, the restless craving for rotation in office, the insistence that the prizes be widely distributed, would make it highly improbable that a senator would be given more than one or, at most, two terms. When the loss to the country is estimated if the service of a Webster or a Clay, a Sherman or a Hoar, were limited to six or even to twelve years, the innovator may well hesitate to urge popular election; for the evidence is incontrovertible that the American people still cherish the notion of rotation of office, and that they are particularly loath to reelect men for long terms of legislative service.

The extent to which Haynes’ prediction has been shown to be incorrect is almost laughable. And so, indeed, were certain of his suppositions.

    It was Haynes’ opinion – based on Lord knows what data – that allowing Senators to be chosen by the various state legislators produced, “A continuity of service, and hence an efficiency based upon long experience in legislature work.” In the event that the general electorate were instead granted the responsibility, “The trading of localities,” he said, “The restless craving for rotation in office, the insistence that the prizes be widely distributed, would make it highly improbable that a senator would be given more than one or, at most, two terms.” In actual fact, however, rather the opposite was demonstrably the case. “When the loss to the country is estimated [,]” he said, “If the service of a Webster or a Clay, a Sherman or a Hoar, were limited to six even to twelve years, the innovator may well hesitate to urge popular election [.]” But long did Daniel Webster actually serve? Three terms, all told. And Henry Clay? Two and a half. And John Sherman? Three and a half. And George Hoar? Four and a half. Longer tenures, to be sure, than Haynes feared would become the norm, but by how much in real terms? Would Henry Clay’s career have been so much the worse for the lack of three additional years in the Senate? Or Webster’s for the lack of six? And what of all the other men who served far shorter stints alongside these notable aberrations? Were they not the norm in both notoriety and tenure in office? Would it have been so unforgivable if fewer individual Senators came to dominate the chamber’s discourse for upwards of a dozen years on end? Would it not have benefited the American people for influence within the Senate to be distributed in a somewhat more equitable manner?

    In actual fact, as aforementioned, Haynes’ prognostication could not have been more wrong. While the relevant trend was somewhat slow to develop, Senate tenures, on average, have only lengthened since 1913. Consider, by way of example, the periods of service rendered by certain Senators in the states of New York, Virginia, and Pennsylvania both before and after the adoption of the 17^th Amendment. Prior to 1913, the longest tenure served by a Senator representing the state of New York was that of Republican Roscoe Conkling (1829-1888), who was elected to three terms between 1867 and 1881, the last of which he voluntarily cut short. In Virginia, this same honor goes to Democrat John W. Daniel (1842-1910), who was elected to an impressive five terms between 1887 and 1910. And in Pennsylvania, the longest pre-reform tenure was that of Republican J. Donald Cameron (1833-1918), who was elected to four terms in the twenty years between 1877 and 1897. These men were not representative of the norm, to be sure; most Senators managed only one or two terms before resigning in favor of a federal appointment, retiring from public life altogether, or dying in office. But they do, between them, demonstrate that lengthy periods of service in the Senate were actually possible under the original, unamended terms of the United States Constitution. Compared to their successors in the post-amendment era, however, their sustained careers in the upper chamber take on a somewhat more mundane appearance.

    In New York, for example, since the ratification of the 17^th Amendment, Conkling’s former record of three terms has been beaten several times over. In fact, the race for longest tenure representing the Empire State in the Senate is at present a four-way tie between Robert F. Wagner (1877-1953), Jacob Javits (1904-1986), Daniel Patrick Moynihan (1927-2003), and the currently serving Charles Schumer (1950-present), all of whom, over the course of their careers, managed four successful elections. In Virginia, Daniel’s previously exceptional five-term stint has likewise been obliterated by that of Democrat Harry F. Byrd (1887-1966), who racked up a total of seven terms between 1933 and 1965. And finally, in Pennsylvania, Cameron’s four-term watermark has since been bested by Republican-turned-Democrat Arlen Specter (1930-2012), who won a total of five elections between 1980 and 2011. The average Senate tenure has likewise been significantly lengthened. Whereas, prior to 1913, only fourteen Senators from the three states under discussion were elected to three terms of more, since 1913, that number has increased to twenty. Single-term Senators have also become especially rare, while two terms in office have become more or less the norm. Particularly competitive states like Pennsylvania and Virginia do rather throw this average off, it must be said, with both states having experienced a string of single-term Senators at some point in the last sixty years. But by and large, when one considers these statistics in the light of Haynes’ aforementioned forecast, the depth of his miscalculation is nonetheless readily apparent. Far from shortening the average Senate tenure, the century that followed the ratification of the 17^th Amendment instead witnessed a general lengthening of service. Not only have a greater number of individuals been elected to two terms in office or more, but the inhabitants of the states have shown themselves more amenable to electing a given Senator three times or more than the members of the various state assemblies ever seemed to be.