Repeal, Reform and the Debate Today

This last lesson brings the course to the present. Lesson 8 and lesson 9 set out the argument over what the amendment did. Here the question is what, if anything, people want to do about it: who has proposed repeal or reform, what the states have actually voted on, what stands in the way, and what each side says at its best. The lesson ends with the course's figures, assembled from primary sources so that any reader can check them.

Where the repeal idea came from

For most of the twentieth century the Seventeenth Amendment was treated as settled and uncontroversial. The scholarly challenge came in the 1990s, alongside the Supreme Court's federalism revival and the term limits movement. Todd Zywicki's articles in the Oregon Law Review (1994) and the Cleveland State Law Review (1997), C. H. Hoebeke's The Road to Mass Democracy (1995), Jay Bybee's "Ulysses at the Mast" (1997) and Ralph Rossum's Federalism, the Supreme Court, and the Seventeenth Amendment (2001) are the works that made the amendment an argument again. None of them was a campaign document. Zywicki's 1997 article is about how to judge other reforms in light of the amendment; Bybee doubts repeal is realistic; Rossum's target is the Court. But together they supplied a vocabulary, "the states' ambassadors," "the lost structural check," "the irony," that politicians picked up a decade later.

The 2010 season

The idea moved from the law reviews to the campaign trail in 2010. The sequence, from dated sources:

Date	Who	What	Source
September 16, 2010	Randy E. Barnett (Georgetown law professor) and William J. Howell (Speaker of the Virginia House of Delegates)	Wall Street Journal op-ed, "The Case for a 'Repeal Amendment'": an amendment letting two-thirds of state legislatures repeal any federal law or regulation. The piece says the Seventeenth Amendment removed a check that legislatures once provided, and concedes the proposal "will not cure all the current problems with federal power."	WSJ (Cato copy)
November 10, 2010	Todd Zywicki	National Review, "Repeal the Seventeenth Amendment." The article's summary notes that Joe Miller, the Republican Senate nominee in Alaska, had backed repeal, an idea "rapidly gaining steam in Tea Party circles," and later backed away.	National Review
November 12, 2010	David H. Gans, Constitutional Accountability Center	Reply to Zywicki: the reform came from the states themselves, 27 of which had applied for a convention by 1910; indirect election bred corruption and deadlock; the federalism objections were heard in Congress and rejected.	CAC
November 2010	Justice Antonin Scalia	At a Texas law school appearance, said he favored returning to the pre-1913 method: "We changed that in a burst of progressivism in 1913."	Time, Nov. 24, 2010
2010	Mike Lee, newly elected senator from Utah; Ken Buck, Senate nominee in Colorado	Time reported Lee had called the amendment's passage a "mistake" and that Buck had backed repeal before opposing it.	Time
February 2011	Texas Representative Truitt (author)	HCR 19 (82nd Legislature) memorializing Congress to propose the Repeal Amendment. Its preamble says the Sixteenth and Seventeenth Amendments "greatly expanded the power of the federal government at the expense of the states." Filed November 8, 2010; referred to the Select Committee on State Sovereignty February 18, 2011; no further action.	Texas Legislature Online

Two features of that season matter for what followed. First, the most concrete proposal, Barnett and Howell's, was a state veto rather than a return to legislative election; it used the Seventeenth Amendment as a diagnosis and prescribed something else. Second, candidates who raised repeal tended to retreat from it, which is an early sign of the polling problem discussed below.

The states act: 2002 to 2025

State legislatures have gone both directions. Three ratified the amendment long after it took effect. One asked Congress to repeal it. One chamber in another state agreed, and the other chamber refused.

State	Measure	Action	Date	Source
Alabama	Ratification of the Seventeenth Amendment	Ratified (symbolic; the amendment had been in force since 1913)	April 16, 2002	House Manual, 117th Congress
Delaware	Senate Concurrent Resolution 30, 145th General Assembly (Sen. Sokola; Rep. B. Short)	Passed both houses; "ratifies the 17th Amendment to the U. S. Constitution, requiring popular vote for U. S. Senators"	July 1, 2010	Delaware General Assembly
Texas	HCR 19, 82nd Legislature (Rep. Truitt): memorial for the Repeal Amendment (state veto)	Referred to committee; died	February 18, 2011	Texas Legislature Online
Maryland	House Joint Resolution 3, enacted as Joint Resolution 4 of 2012	Ratified; the resolution recites that Maryland had never acted and "now wishes formally to record its support"	Signed by the presiding officers May 22, 2012	Maryland General Assembly
Utah	SJR 2, 2016 General Session (Sen. Alvin B. Jackson; Rep. Ken Ivory): Joint Resolution Calling for the Repeal of the 17th Amendment	Passed Senate 20 to 6 with 3 absent (February 24, 2016); passed House 39 to 34 with 2 absent (March 2, 2016); filed with the Lieutenant Governor March 15, 2016	March 2, 2016	Utah Legislature
Montana	SJ 22, 69th Legislature (Sen. Greg Hertz): Joint resolution to repeal the 17th amendment to the US constitution	Passed Senate on third reading 29 to 19 (April 5, 2025); House refused to concur 35 to 64 (April 24, 2025)	April 24, 2025	Montana Free Press Capitol Tracker; official bill page
Idaho	See lesson 7

Utah's resolution deserves a closer look because it is the only one that passed and because it supplies a proposed text. Its preamble argues that the founders meant legislative authority to be divided between two chambers with different modes of election, cites Madison's Federalist Nos. 10 and 62, and claims that popular election diluted state power and increased federal power over the states. Its resolving clause urges Utah's delegation and every member of Congress to "propose an amendment to the United States Constitution repealing the Seventeenth Amendment." The proposed text would restore legislative selection with six-year terms, give the legislature a power to remove the senator it selected, and create liaison committees between the legislature and its senators. That last pair of features answers one of the counter-case's points from lesson 9: a bare return to the old text would restore legislatures that could not control senators once seated, so the drafters added a removal power.

Montana's 2025 vote shows the ceiling. A repeal resolution passed one chamber comfortably and lost the other by nearly two to one, with the sponsor's own party split 35 to 22 in the House, according to the Montana Free Press tally of the recorded votes. (The official Montana bill page is linked above; its vote records load in a browser application rather than a static page, which is why the tracker is cited for the tallies.)

The 2020 revival

On September 8, 2020, Senator Ben Sasse of Nebraska published a Wall Street Journal op-ed proposing a set of Senate reforms that included repealing the Seventeenth Amendment, as National Review reported the next day. The op-ed itself is behind the Journal's paywall and could not be retrieved for this lesson, so his exact words are not quoted here. The proposal drew replies from several directions within days, including from writers who agreed with Sasse about the Senate's dysfunction and disagreed about the remedy. It did not produce legislation. Its significance is that a sitting senator, who had won two statewide elections under the amendment, put repeal on the table, which had not happened in the 2010 season.

Reform proposals short of repeal

Most of the serious thinking since the 1990s has gone into alternatives. Four have identifiable authors and texts.

Proposal	Author and source	What it would do	What it would not do
The Repeal Amendment (state veto)	Randy Barnett and William Howell, WSJ, September 16, 2010; Texas HCR 19 (2011)	Let the legislatures of two-thirds of the states repeal any federal law or regulation by resolutions describing it; Congress could reenact by simple majority	Change how senators are chosen; prevent new federal laws; work without 34 legislatures agreeing on a specific provision
Legislative selection with a removal power and liaison committees	Utah SJR 2 (2016), proposed amendment text	Return selection to legislatures, add a power to remove the senator, and formalize communication between legislature and senator	Address deadlocks, money in legislative races, or the national-party problem Schleicher describes
Judicial restraint on federalism	Ralph Rossum, Federalism, the Supreme Court, and the Seventeenth Amendment (2001)	Have the Court stop substituting doctrines such as anti-commandeering for the structural protection the people removed in 1913	Restore any state check; it accepts the amendment and asks the Court to accept it too
Election law reform for state democracy	David Schleicher, 65 Hastings L.J. 1043 (2014)	Change election rules so that state legislative elections turn on state performance rather than the president's popularity	Touch the Senate at all; Schleicher argues repeal would make legislatures "electoral colleges for U.S. Senators"

Other ideas circulate without a verified author: a hybrid in which legislatures nominate and voters choose among the nominees, or a legislative power to recall senators without repealing direct election. They are described here only as ideas discussed; no published proposal with a named author and date could be confirmed this month, and readers should treat attributions of them to particular scholars with caution. Zywicki's 1997 article is the best single source on how reform proposals of the 1990s, term limits among them, look when measured against the structure the amendment removed.

The obstacles

Article V

Article V gives two routes. Congress proposes an amendment when "two thirds of both Houses shall deem it necessary," or calls a convention "on the Application of the Legislatures of two thirds of the several States." Either way the amendment takes effect "when ratified by the Legislatures of three fourths of the several States" or by conventions in three-fourths. With 100 senators, 435 representatives and 50 states, that is 67, 290 and 38. The first route asks 67 senators to vote to end the elections that put them there. The second route has never been used. The convention applications of 1893 to 1912 (lesson 4) came within a few states of two-thirds and forced the Senate's hand without a convention being called; no comparable campaign for repeal exists.

Public opinion

Only one national poll on the question could be located this month on a pollster's own site. YouGov, with the Huffington Post, surveyed 1,000 U.S. adults on November 4 and 5, 2013 (margin of error 4.7 points) and published the results on November 6. Asked which method is better, 71 percent chose electing senators directly by popular vote, 11 percent chose having state legislatures elect them, and 18 percent were not sure. Asked whether the Seventeenth Amendment should be repealed, 16 percent said yes, 64 percent said it should remain, and 21 percent were not sure. The poll is twelve years old and one survey is thin evidence, but the margins are so wide that a reversal would be remarkable. Other polls may exist; none was found on a pollster's own page, and secondhand summaries are not used here.

The late ratifications

Alabama in 2002, Delaware in 2010 and Maryland in 2012 went to the trouble of ratifying an amendment that had been in force for nine decades. Delaware's resolution came in the same summer the repeal idea was gaining ground in Senate campaigns; Maryland's followed the Utah senator's remarks and the Texas resolution. These acts have no legal effect. They are legislatures voting, on the record, that they do not want the power back.

What repeal would and would not do

Suppose the obstacles vanished and the Seventeenth Amendment were repealed tomorrow, restoring Article I, Section 3 as written. What would follow is better known than the debate suggests, because it happened once.

• Legislative elections would become Senate elections. Schleicher's point, drawn from the 1890s and 1900s: when a legislature's main national act is choosing a senator, voters choose legislators for their Senate preference, and state government is judged on national politics. National parties are stronger now than in 1900. Repeal would make state legislatures, in his phrase, electoral colleges for senators.
• Money would move to legislative races. Schiller and Stewart's estimate for the indirect era was about $12.64 million per Senate election in 2012 dollars, with Senate candidates recruiting and funding legislative candidates. Campaign finance law today regulates candidates and committees; it has no category for a Senate aspirant financing fifty legislative races to build a majority. The dark money lesson in the campaign finance course describes how hard the current system is to trace; a repeal would add a layer.
• Deadlocks and vacancies would return unless the design changed. Haynes's record (below) shows 46 deadlock entries and 14 failures to elect in fifteen years under the 1866 Act's daily-ballot rule. Any modern repeal would need a tie-breaking rule, and the Utah text does not supply one.
• The removal power would be new. Utah's proposed text would let legislatures remove senators. Nothing like it existed before 1913; instruction was a norm, not a power. This would be a stronger state check than the framers wrote, and it would make senators' six-year independence conditional on a legislative majority that can change every two years.
• Federal spending would not fall by itself. Lesson 9's OMB tables show growth driven by wars, the Depression and entitlement programs. A legislature-chosen Senate would face the same voters' demand for Medicaid, highways and disaster aid, routed through legislators who want those things for their districts.
• Equal suffrage would be unchanged. The smaller states' weight in the Senate, protected by Article V, is the same with or without the amendment.

What repeal would do, on its proponents' own account, is change the senator's principal and therefore the senator's incentives at the margin: a senator who could be removed by a legislature would think twice about voting for an unfunded mandate or a preemption clause the legislature opposed. The structural argument of lesson 8 is an argument about margins over decades. Its opponents reply that the margin would be small and the costs certain.

The arguments, each at its strongest

For repeal or a legislative role	For keeping direct election
The Constitution protected the states by structure, and Madison said in Federalist 62 that legislative election was the link between the two systems. Removing a structural check and hoping courts fill the gap has produced a Court that lurches between Garcia (1985) and Printz (1997) by one vote.	The people amended the Constitution by the method it provides, after twenty years of debate, with 36 states in eleven months. Rossum's own irony cuts here: a constitutional choice made that deliberately deserves respect, and the Court's job is to apply the amended structure.
A senator who can be removed by a legislature, as Utah's text proposes, has a concrete reason to oppose unfunded mandates, coercive grant conditions and preemption clauses. No such reason exists today, which is why Congress needed UMRA in 1995 and the states needed NFIB in 2012.	State governments win in Congress constantly through governors, attorneys general and associations; UMRA itself was passed by a directly elected Congress. The behavioral studies show direct election made senators more responsive to voters, which is what a republic wants its legislators to be.
The old system's defects were fixable. Deadlocks came from the 1866 Act's rules; bribery was prosecuted and punished (the Senate declared Lorimer's election invalid in 1912, 55 to 28); a modern design could add tie-breakers and disclosure. The Progressives removed the structure to cure its symptoms.	Schiller and Stewart's 752 elections show the defects were the system: party caucuses, money and ambition decided outcomes 98 percent of the time along party lines, and legislators who served one term could not supervise anyone. Repeal restores 1899, not 1789.
Interest groups now lobby one kind of constituency twice. Restoring a different constituency for the Senate would restore the bicameral filter Zywicki describes, and the states are the natural second constituency.	National parties would make legislatures the filter in name only, as Schleicher argues, and state elections would again be decided by national mood. The cost would fall on state democracy, the thing the amendment's opponents say they want to protect.
Public opinion is uninformed on the question (a fifth "not sure" in 2013) and movable, as the 2016 Utah vote and the 2025 Montana Senate vote show.	Public opinion is lopsided (71 to 11 on method, 64 to 16 on repeal) and the only state votes that passed were memorials with no legal effect; three states ratified late to say the opposite.

Questions and answers

Has Congress ever voted on repealing the Seventeenth Amendment?

No vote on such a proposal was found in this course's research. The state memorials in Utah (2016) and the Montana Senate (2025) ask Congress to act; Texas's 2011 resolution asked for a different amendment. Senator Sasse's 2020 op-ed proposed repeal but was not followed by a resolution that reached a vote. Anyone claiming a congressional vote should be asked for the Congressional Record citation.

Could the states repeal it without Congress?

Only through an Article V convention called on the application of two-thirds of the state legislatures (34), followed by ratification by three-fourths (38). Article V has never produced a convention. The campaign for direct election itself came close to the two-thirds threshold between 1893 and 1912 (lesson 4) and the Senate acted to avoid one.

Why would a state ratify an amendment that was already in force?

To put itself on record. Maryland's Joint Resolution 4 recites that the General Assembly had never acted and wished to record its support formally by ratifying. Delaware's SCR 30 is a one-sentence ratification. The timing, 2010 and 2012, matches the years repeal was being discussed in Senate campaigns.

What would Utah's proposed amendment actually say?

The enrolled SJR 2 supplies text restoring legislative selection with six-year terms, adding a power for the legislature to remove the senator, and creating liaison committees. It asks Congress to propose that text. It is a memorial; it does not itself amend anything.

Did Justice Scalia say the amendment should be repealed?

Time's Adam Cohen reported on November 24, 2010 that at a Texas law school appearance that month Scalia said he favored the pre-1913 method and described the 1913 change as a burst of progressivism. That is a remark at a public event as reported by a magazine, not a judicial opinion, and it should be described that way.

Is there any evidence that repeal would reduce federal spending?

None that this course could find. The structural argument predicts an effect at the margin over time; the OMB tables in lesson 9 show spending driven by wars and entitlements that voters of every state want. The honest statement is that the claim is theoretical and the counter-evidence is historical.

How many senators have been appointed rather than elected since 1913?

The Senate Historical Office's list runs to 210 appointments from January 1913 through July 2026, with the decade-by-decade counts in the table below. Of the 203 whose outcome is recorded, 70 later won election to the seat, 62 ran and lost (in a primary or general), and 71 did not seek election. Lesson 6 covers the vacancy clause and the court cases; lesson 7 covers Idaho's six appointments.

Where can I read the strongest case against repeal from someone who shares the federalism concern?

Schleicher's 2014 article. He accepts that federalism matters and argues that repeal would damage it by nationalizing state elections. For the strongest case for repeal, Zywicki's 1997 article and his 2010 National Review piece are the clearest statements.

Hard questions

If 64 percent oppose repeal and only 16 percent support it, why spend a lesson on a dead idea?

Because legislatures keep voting on it (Utah 2016, Montana 2025), a sitting senator proposed it in 2020, and the argument behind it, about how the Constitution protects the states, is live in the Supreme Court every term. The polling makes repeal unlikely. It does not make the debate over what the amendment did any less important, and a reader who understands that debate can judge the next proposal on its merits.

Isn't it anti-democratic to even consider taking the vote away from people?

The framers thought a republic needed filters, and the Senate was one of them; the people of 1913 decided otherwise by the amending process; the people of 2013 told YouGov they agreed by 71 to 11. Proponents of repeal answer that representative democracy already works through intermediaries (the Electoral College, the legislatures that ratify amendments) and that the question is which intermediary produces better government. That is a legitimate constitutional argument, and the way to meet it is with the evidence in lessons 8 and 9 rather than by declaring it out of bounds.

Utah's resolution passed. Doesn't that show real momentum in the states?

It shows one legislature, in the state that rejected the amendment in 1913, passing a memorial by 39 to 34 in its House. The next legislative test, Montana in 2025, failed 35 to 64 in the House with the sponsor's party split. Three other states ratified the amendment late to say the opposite. Momentum would look like Article V applications accumulating, as they did from 1893 to 1912. Nothing like that exists.

The reform proposals short of repeal are either toothless (Rossum, Schleicher) or need an amendment anyway (the Repeal Amendment, Utah's text). So the only real choice is repeal or nothing.

Two of the four need an amendment, two do not, and the two that do not are the ones that address the actual mechanisms. Rossum's proposal changes what the Court does, which is where the federalism fights of 1992 to 2012 were decided. Schleicher's changes state election law, which is within each state's power now. Whether they are "toothless" depends on whether you think the problem is in the Senate's incentives or in the Court and the states' own politics. Lesson 9's scorecard suggests the second.

Your appointments table shows 210 senators who never faced voters before taking office. Isn't that already a partial repeal of direct election, run by governors?

It is the amendment's own proviso at work: legislatures may let governors fill vacancies "until the people fill the vacancies by election." The count is real and so is the pattern that appointees often do not run or lose (62 defeated, 71 did not run, out of 203 with a recorded outcome). The remedy for a reader who dislikes it is in the state's vacancy law, which the legislature can change at any time to require a quick special election, as lesson 6 explains some states do. That is a far easier reform than any amendment, and it is one the states hold entirely in their own hands.

Pictures, Numbers and Infographics

The timeline runs from the Constitutional Convention to the latest state vote. Every date on it appears in one of the lessons of this course with its source. Below it are three tables: the ratification dates from the Constitution Annotated, the deadlocks of 1891 to 1905 from Haynes's 1906 record and the Senate Historical Office, and the Senate's own list of appointed senators counted by decade.

Ratification of the Seventeenth Amendment

Dates are from the ratification note to Amendment XVII in the 2016 Constitution Annotated, which records the proposal on May 13, 1912 (37 Stat. 646), completion on April 8, 1913 with 48 states in the Union, and Secretary Bryan's certification on May 31, 1913 (38 Stat. 2049). The House Manual for the 117th Congress gives the same list with three differences, noted in the last column. Where the two official compilations differ, both dates are shown rather than one chosen.

Order	State	Date (Constitution Annotated)	Note
1	Massachusetts	May 22, 1912
2	Arizona	June 3, 1912
3	Minnesota	June 10, 1912
4	New York	January 15, 1913
5	Kansas	January 17, 1913
6	Oregon	January 23, 1913
7	North Carolina	January 25, 1913
8	California	January 28, 1913
9	Michigan	January 28, 1913
10	Iowa	January 30, 1913
11	Montana	January 30, 1913
12	Idaho	January 31, 1913	See lesson 7 for the chamber votes
13	West Virginia	February 4, 1913
14	Colorado	February 5, 1913
15	Nevada	February 6, 1913
16	Texas	February 7, 1913
17	Washington	February 7, 1913
18	Wyoming	February 8, 1913
19	Arkansas	February 11, 1913
20	Illinois	February 13, 1913
21	North Dakota	February 14, 1913
22	Wisconsin	February 18, 1913
23	Indiana	February 19, 1913
24	New Hampshire	February 19, 1913
25	Vermont	February 19, 1913
26	South Dakota	February 19, 1913
27	Maine	February 20, 1913	House Manual: February 11, 1913
28	Oklahoma	February 24, 1913
29	Ohio	February 25, 1913
30	Missouri	March 7, 1913
31	New Mexico	March 13, 1913
32	Nebraska	March 14, 1913
33	New Jersey	March 17, 1913
34	Tennessee	April 1, 1913
35	Pennsylvania	April 2, 1913
36	Connecticut	April 8, 1913	Three-fourths of 48 reached
37	Louisiana	June 5, 1914	House Manual: June 11, 1914
38	Alabama	April 16, 2002	House Manual only (the 2016 Constitution Annotated note omits it)
39	Delaware	July 1, 2010	SCR 30, 145th General Assembly; the House Manual records Delaware's rejection on March 18, 1913
40	Maryland	May 22, 2012	House Joint Resolution 3, Joint Resolution 4 of 2012
Rejected	Utah	February 26, 1913	Both compilations; Utah urged repeal in 2016
No action	Florida, Georgia, Kentucky, Mississippi, South Carolina, Virginia, Rhode Island		The House Manual lists Florida, Georgia, Rhode Island and South Carolina as not completing action; Kentucky, Mississippi and Virginia appear on neither list

Deadlocks in Senate elections, 1891 to 1905

From the table "Record of Deadlocks" on pages 38 and 39 of George H. Haynes, The Election of Senators (1906). Haynes defined a deadlock as "a prolonged and stubborn contest," counted days from the two houses' first separate ballots to the final vote, and drew his data from Appleton's Annual Cyclopedia, the New York Tribune Almanac, state journals and correspondence with secretaries of state. Ballot counts marked "c." were his estimates. Haynes's table has 46 entries, because the two-seat contests in Delaware (1901 and 1903) and Nebraska (1901) are listed once per seat, and 14 of the entries end in no election; the Senate Historical Office gives the same figures, 46 and 14. The table below merges each two-seat contest into one row, so it has 43 rows. The scanned text is imperfect in two cells, noted below.

Year	State	Days	Ballots	Result
1891	Florida	35	about 75	Wilkinson Call elected
1891	North Dakota	3	17	H. C. Hansbrough elected
1891	South Dakota	27	40	J. H. Kyle elected
1892	Louisiana	44		No election
1893	Montana	50	44	No election
1893	Nebraska	21	17	W. V. Allen elected
1893	North Dakota	33	61	W. N. Roach elected
1893	Washington	51	101	No election
1893	Wyoming		no ballot possible	No election
1895	Delaware	114	217	No election
1895	Idaho	51	52	G. L. Shoup elected
1895	Oregon	32	58	G. W. McBride elected
1895	Washington	9	28	J. L. Wilson elected
1896	Kentucky	58	52	No election
1896	Louisiana	9	6	S. D. McEnery elected
1896	Maryland	8	7	G. L. Wellington elected
1897	Florida	24	about 45	S. R. Mallory elected
1897	Idaho	15	not legible in the scans	Henry Heitfeld elected
1897	Kentucky	36	60	W. J. Deboe elected (special session)
1897	Oregon	53	no ballot possible	No election
1897	South Dakota	29	27	J. H. Kyle elected
1897	Utah	17	53	J. L. Rawlins elected
1897	Washington	7	25	George Turner elected
1898	Maryland	7	10	L. E. McComas elected
1898	Tennessee	7	7	T. B. Turley elected
1899	California	67	104	No election
1899	Delaware	64	113	No election
1899	Montana	17	17	W. A. Clark elected (see lesson 3)
1899	Nebraska	50	43	M. L. Hayward elected
1899	Pennsylvania	92	79	No election
1899	Utah	52	164	No election
1899	Wisconsin	8	6	J. V. Quarles elected
1901	Delaware (two seats)	52	46	No election
1901	Montana	51	66	Paris Gibson elected
1901	Nebraska (two seats)	72	54	C. H. Dietrich and J. H. Millard elected
1901	Oregon	22	53	J. H. Mitchell elected
1903	Delaware (two seats)	41	36	J. F. Allee and L. H. Ball elected
1903	North Carolina	10	9	L. S. Overman elected
1903	Oregon	32	42	C. W. Fulton elected
1903	Washington	9	13	Levi Ankeny elected
1904	Maryland	16	12	Isidor Rayner elected
1905	Delaware	80	51	No election
1905	Missouri	60	67	William Warner elected

Haynes also listed the seats that stayed empty as a result (pages 62 and 63). "Six States have accepted vacancies in the Senate as the penalty of their legislatures' failure to elect," he wrote, though his table of vacant seats names eight states, two of which (Oregon and California) eventually called special sessions: Montana, Washington and Wyoming from March 4, 1893 to February 1895 (53rd Congress); Delaware from March 4, 1895 to February 5, 1897 (54th); Oregon from March 4, 1897 to December 5, 1898 (55th); Delaware, California, Pennsylvania and Utah for most or all of the 56th Congress from March 4, 1899; both Delaware seats for the whole 57th Congress (March 4, 1901 to March 3, 1903); and one Delaware seat from March 4, 1905. Of seven Congresses, he wrote, only one had its Senate at full strength. The Senate refused in each case to seat governors' recess appointees (Mantle of Montana and Allen of Washington in 1893, Corbett of Oregon in 1897, Quay of Pennsylvania in 1899), which is why the Seventeenth Amendment's proviso for temporary appointments was written as it was.

Senators appointed to fill vacancies, by decade, 1913 to 2026

Counted from the Senate Historical Office's list Appointed Senators (1913-Present), downloaded this month: 210 appointments, grouped by the decade of the appointment's effective date. "Elected" means the Senate's note records a later election to the seat; "defeated" covers defeat in a general election or for nomination; "did not seek" is the Senate's phrase; "pending or not recorded" covers the seven appointees whose outcome the list leaves blank or records in another way: the four most recent (Ohio and Florida, January 2025; Oklahoma, March 2026; South Carolina, July 2026), two who died in office before an election (1922 and 1941) and one who "was not a candidate in special election" (1950s). Idaho's six appointments on the list (1918, 1928, 1940, 1945, 1949 and 1962), and a 1921 appointment the list excludes by rule, are covered in lesson 7.

Decade	Appointments	Later elected	Defeated	Did not seek election	Pending or not recorded
1910s (from January 1913)	15	6	3	6	0
1920s	26	12	8	5	1
1930s	29	11	8	10	0
1940s	39	10	11	17	1
1950s	21	3	7	10	1
1960s	20	7	10	3	0
1970s	11	1	6	4	0
1980s	6	3	2	1	0
1990s	9	3	3	3	0
2000s	13	7	1	5	0
2010s	12	5	2	5	0
2020s (through July 2026)	9	2	1	2	4
Total	210	70	62	71	7

The 1940s had more appointments than any other decade, 39, and the 1980s the fewest, 6. The share of appointees who went on to win the seat has never reached half in any decade, which bears on the argument in lesson 6 about how long an appointee should serve before the people choose.