Researcher Talk: Electing the Senate: Indirect Democracy Before the Seventeenth Amendment
Articles,  Blog

Researcher Talk: Electing the Senate: Indirect Democracy Before the Seventeenth Amendment


Richard McCulley: Thank you for attending
today’s Researcher Talk. I am Richard McCulley, the Historian at the
Center for Legislative Archives that sponsors this series. Today’s talk closes out a really superb series
of presentations that we’ve heard this year. We will hibernate in December and January
and come storming back in February with monthly programs in 2016. We are delighted to host Charles Stewart,
the Kenan Sahin Distinguished Professor of Political Science at MIT where he has taught
since 1985. He is a Fellow at the American Academy of
Arts and Sciences. Charles is also a longtime friend and supporter
of the Center for Legislative Archives. When the Center was created and the Congressional
Research Room opened with full runs of the printed serial set, the Congressional Record,
and its predecessor publications, the Senate and House journals, probably no researcher
spent more time working in that room than Charles. He was busy compiling the membership of Senate
and House committees for the multi-volume Committees in the United States Congress,
which is an absolutely indispensable source of congressional scholars. Back in this pre-digital era, this work had
to be done the old fashion way and it was all there in one place. Charles regarded the Congressional Research
Room as a marvel, declaring it one of his favorite places in the entire world. A couple of years ago when we heard of Charles’s
and Wendy Schiller’s project on Senate elections before the 17th Amendment, I tried to get
him to speak about that work-in-progress. Charles wisely held off and early this year,
we schedule today’s talk. Princeton University Press published the book
earlier this year and it has already been nominated for at least one book award that
I am aware of. But in September and October, our plans to
talk today about the Senate elections nearly got derailed due to all of the excitement
and chaos in the selection of a new speaker of the House. Two years ago, Charles and Jeff Jenkins published
Righting for the Speakership: The House and the Rise of Party Government, also published
by Princeton University Press. The publicity train had just left the station
concerning today’s talk on the Senate. We decided the issue of how the speaker is
selected was one that would not be resolved soon and Charles has agreed to return next
year and give a talk on that timely topic. It will certainly be a not-to-be missed Researcher
Talk in 2016. We should have time for Q & A. Please raise
your hand if you have a question so that we can pass the microphone. Thank you for joining us today, Charles. Stewart: Thank you, Richard, for that very
kind introduction. So the word “marvel” was used and you’ve already
trumped a little of my throat-clearing but it is a thrill for me to be giving a talk
here today. I was thinking coming over here today. I arrived a little early and was thinking
that I wrote my dissertation in 1983-1985. I came to Brookings to write it and I was
writing about the history of the appropriations process and budget reform. That book ended up going from 1865 to 1921
and I spent almost the entire two years here if I wasn’t up at the Library of Congress. Basically, the research was all done here. And I remember the days of taking the Yellow
Line to the Archives stop. And in those days you would not even want
to stand at the corner waiting for the light to turn. So I would just get out a high-tail it straight
across the street to get into entrance and come up to the third floor, fifth floor to
the big, main reading room there. Now the fact that I can come in and have a
croissant across the street and saunter across the way. It is certainly a much more diverse place
and more interesting neighborhood than it used to be and safer. Not too many political scientists, almost
none of them back thirty years ago did historical work, especially a quantitative historian
like I did. So to learn about the treasures here and Charles
South, some of you may remember, who I worked with at that time was indispensable in teaching
me about the evolution of the House of Representatives. In the last thirty years there has been a
growth in interest in congressional history. I have been part of that growth, but without
the folks there then and over the last several years, I do not think what I do would have
proceeded as it has. So I am going to jump into the talk. I am a political scientist – I mean there
are stereotypes of historians and political scientists. And I am tempted to break into a rendition
of “The farmer and the Cowman [should be friends” from Oklahoma, one of my favorite moments
in American cinema. I would just say that the thing I particularly
appreciate about historians that I am not particularly good at… and I am spinning
a good tale – although one of the reasons that I will mention that this is a great project
is because of the great tales that one can spin as a political scientist. We take the richness of history and batten
it and beat it around the head and shoulders until it fits into a computer and then we
do graphs. I will show you some of that. And give you a sense of some of the really
interesting stories and episodes that go along with this project. But I am a political scientist so I will stick
to what I know. Here is the book cover. And I didn’t even realize until my co-author,
Wendy Schiller, procured the rights to the cover, but I realize now that the cover actually
comes from the Clifford Berryman Collection of political cartoons from the National Archives. That’s the cover, by the way -.
Here is the table of contents for the book. It is too small to read so I’ll note that
what Wendy and I do in the book – we look at senate election in the state legislature
from 1871 to 1913. We have this standard data collection. But we also have a number of themes that we
address, reflected in the table of contents, such as the theme of candidate emergence,
the dynamics of representation, and the effect of direct/indirect elections in the teens. I am not going to follow that logic of the
book, in part because I thought given this crowd in the audience today, that it might
be more useful and helpful to get into the project by talking about some of the larger
issues, the sources and where material like this comes from. One of the enduring issues that comes out
of our research having to do with the role of political parties and elections of senators
during this period – and then throw it open to discussion. And then if we have enough time, I will interpret
the book cover which is kind of an interesting book cover. If it is okay with the technical people, I
am going to stand a little bit just because of the setup here. I want to say a couple of words about why
this issue is important and interesting. We got into it because of the importance of
the Constitution which is the enduring touchstone of American politics. But also, political scientist, historians,
and also the general public has an interest – It is one of the areas where the interests
of the academics and the public comes together – in the topic of what I call the – “Founders’
Constitution” on the document – how it was written and implemented in the early years
and how that Constitution interacted with changes in American society over the next
century. Most important is the rise of political parties
under Andrew Jackson, actually under Martin Van Buren who is one of my “heroes” in American
political history. Then in the late nineteenth century – and
I think that it is important to the story that we tell – the encounter of the Founders’
Constitution with the rise of industrialization, large complex organizations, large political
parties, large business enterprises, etc. I think that there is something importance
to both the public and to scholars and to the chattering classes of whom there are many
in Washington D.C. It is important for many reasons. The first two that I’ve noted here are a little
more inside baseball sorts of things but we’ll see that the form of the state legislatures
is common in American politics but not really very well-understood. By which I mean that the election of senators
during the indirect election period – how within the state legislature you had to get
a majority of all of those cast and there was no natural method to narrow down to just
two candidates. So you could conceivably cycle for a long,
long time before coming up with a winner who got a majority. And this is the same procedure that remains
in the Electoral College provision that remains in the Constitution in case there is a tie
or no one gets a majority. So it is still actually in the Constitution
– And political conventions. If we go into next summer without a clear
majority winner on the Republican side, this is the form of election that is going to follow. It’s important. It is more common than we think and we don’t
understand much about the formal or historical nature of these things. So common, but understudied. And as I mention here, there are stabbings,
militias, thugs, prosecutorial misconduct – And that is just in one election. The Kentucky election of 1897 which I hope
to say a few words about. Sometimes it is a snoozer, but sometimes it
has everything you can imagine in late-19th century American politics, especially in Frankfurt,
Kentucky – A little about previous scholarship and I
will not dwell on this. There has been a small amount of scholarship
about senate elections before direct elections. George Haynes is probably the major source
for discussion about these elections. He wrote two books. One was about election of senators in 1908
and then his larger book about the Senate in general. Two things that I can say about this. It is typical “Progressive History” in the
sense that George Haynes really didn’t like indirect election. He also cherry picked. Woodrow Wilson’s Congressional Government
is probably the best-known example of this genre. They had very strong ideas about what government
should look like and how it should proceed and an aversion during this period – but if
you dig in deeply, not a lot of attention to the empirical bases of the situation. A lot of times they were right, but we just
don’t know empirically. So that’s George Haynes. Even Rothman’s Politics and Power, one of
the best-known histories of the Senate for this period. And there are some accounts of elections in
that book. It is a very good book, written in 1966, for
getting a sense of what it was like to be a senator in the Gilded Age. And then Bill Riker on the political science
side may not be as much known in this audience, but he is one of the modern founders of American
political science. He was very interested in the Constitution. He wrote an article in 1955 which appeared
in the American Political Science Review entitled “The Senate and American Federalism” and the
empirical part of that article is trying to trace out the path of the Senate being an
agent of state governments to the latter part of the 19th and into the 20th century when
senators really became independent agents and how that happened. He traces out the demise of “instructing”
senators and that sort of thing. He makes a claim that I’ll come back to in
a bit. By 1860, Senate elections were basically popular
elections. And he makes this argument empirically by
taking the Lincoln-Douglas debates – And I always like to ask my students: The Lincoln-Douglas
debates of 1858. They are running for the Senate. And they are: Yes. Yes. Yes. This is the period of indirect elections,
so why are they “running” for the Senate? “Oh.” That’s kind of interesting – They were trying
to drum up support for their political parties in the legislative elections: If you believe
in my principles, then vote for this guy – The big, tall skinny guy – And that was what the
debate was about. Riker generalized about that saying that was
the typical Senate election. And spoiler alert, Bill Riker, as they say,
was more interesting when he was wrong than most political scientists are when they are
right. And he was really interesting here. And finally there is some renewed interest
in this topic among conservative legal scholars – Ted Cruz favors repealing the 17th Amendment
and there is an interest in that. There is not extant data really, although
I notice that someone has begun to populate Wikipedia pages about some of these elections
and I will just skip over that because it just shows some of the problems with Wikipedia. I mean, great effort, but there are some fundamental
problems. But there is some data beginning to appear. The Tribune Almanac which is a wonderful source. I used to read it over dinner sometimes. It is really great from around the 1860s to
about 1910 or so. You can sometimes find in the Tribune Almanac
some things about the elections – Here is an account about the elections in Georgia,
Iowa, Kentucky, Louisiana – little paragraph blurbs. There was a multi-ballot race in Iowa and
they actually have a cable here of the number of candidates and the votes they received. So that is kind of nice. There is actually some data there that we
“seeded” our project with. There is also Appleton’s Annual Encyclopedia
which is another wonderful source, although as an aside concerning what I am primarily
interested in, it is sort of a plagiarism of local newspapers which are kind of cut
and stuck in there, but nonetheless it is a place to start. So this is Appleton’s about Delaware, and
I know you can’t read this, in 1895. One of the crazier elections in Delaware in
which the governor, who had been a state legislator – There was a tie in the state legislature,
so he tried to delay his inauguration as governor so that he could vote for the senator which
cause all kinds of craziness and eventually no one was elected senator. So that is kind of the background intellectual
and data/empirical background to this issue of indirect election. Let me tell you just a little bit about how
senators got elected during this period. The core is the Constitution. Article I that says senators shall be chosen
by the state legislature and they would decide, but that Congress could step in and pre-empt
state legislature which they ended up doing in the case of the Senate. There were some problems with Senate elections
before the Civil War—questions about electoral requirements, quorums – More important things
came along in 1860 or so and came back and in the act of July 12, 1866 to declare a way
in which states would have to elect state senators. And being from MIT, I will describe it as
a flow chart. So there would be a state election. Let’s imagine a November election in an even
numbered year. The state legislature would convene the first
Monday of odd numbered years, often times on New Year’s Day. On the second Tuesday after convening, the
law required that there be a second bicameral election for a senator if there were a vacancy
upcoming. So the vacancy would be filled with the March
3rd session. So we are voting in January for a seat that
is going to come open in March. The next day the two chambers assembly in
a joint session to canvass the votes. They see what the two chambers did -. Of course,
everybody knows what had happened. The question: Does the same person get a majority
in each chamber? If so, you’ve got a senator. If no, and there are a lot of reasons – You
might have two different people getting majorities – there are a lot of reasons, but if the answer
is no, then you would immediately go to the joint assembly voting – Senate and House voting
together with the same majority requirement – and you might get a majority winner. If you do, then you get a senator, but if
you don’t, then you just rinse and repeat. With an important note here that it is the
last day of the session. The requirement is that you vote at least
once every day. If it is the last day of the session, a deadlock
and no senator, and you forego having a senator for the next two years. This happened 19 times, I believe, during
the period that we are talking about. Delaware went without any Senate representation
in the 1890s for a four year period. And this is one of the issues that came up
around the 17th Amendment. Some data sources. Just a few things -. How do we learn about
these elections? The main source of information are the state
journals. And these are the various places that we went. This tells you something about Wendy and me,
the differences between us. Some of these places are the state archives. Wendy went to the University of Miami to get
information about Florida, rather than Tallahassee which is where I would have gone. I went to Frankfurt, Kentucky. She went to New York City. I went to Albany. We tend to like to hang out in different sorts
of places. Some of these are state libraries. Some of these are big institutions. There are four that I will mention, three
of which are particularly valuable for state legislative sources. One is the Library of Congress which has a
nice, big collection. The Yale Law Library has a tremendous collection
of state legislative materials that they have been trying to digitize which is how I found
out about their collection. And also, the New York State Library has a
wonderful collection. They were one of the first to put the catalog
online which allowed me to look at things from afar. But the winner in all of this is the Wisconsin
Historical Society. Not only do I show the Wisconsin Historical
Society, I show the most wonderful thing to a researcher – an open stack. Yes! It’s an open stack and this tells you something
about Wisconsin. Upon admission to the Union before the Civil
War, they went back and collected state legislative documents up to admission and from that point
forward, they continued to collect state legislative documents. They have a full run of every state legislative
journal from the founding of the country. And it is open stack, at least as of a couple
of years ago when I was there. My favorite collection there is the Proceedings
of the Wisconsin Cheese Board which take up an entire range -. We have to have viva voce
voting – everybody has seen in the legislative journals. This is from the New York House of Representatives
Frances Kerin was eventually elected. This is the House and here are the roll call
votes where Kerin wins a majority in the House of Representatives. Oh, by the way, don’t freaked out, the markings
I wrote on the photo copy not on the original. This is the Senate, at the same time in the
same election. Edwin Morgan wins a majority in the Senate. So this is one of the ones where you have
a joint meeting with the proceedings noting that there is no majority in the two chambers. Here are the votes in the joint proceedings
from the joint meeting. Kerin wins and becomes senator. This is typically what we got. Some other sources we used as well, state
blue books, in this case New York red books. Newspapers have a lot of information about
state politics including election returns, I mean amazing stuff. A lot of that gets repeated in the Tribune
Almanac. This is where the fun stories go to die. We ended up creating a a large data base that
records every roll call vote for every senator. It will have information about the election
in the chamber, whether this was a separate or joint ballot and which ballot number. The
candidate, in this case, Francis Brewer voted for Roscoe Conkling. Brewer was from Chautauqua
district one, actually ran as an “administration” party but that was the Republican party, so
we had that information. We have 577,000 roll call records. So 577,000 lines of this. 752
actual elections where we got not only who voted for whom, but things like which party
they belonged to, which district they came from – as much as we could. Still in process,
although the book has been published, there is a lot more than I and Wendy would like
to do to build out the record of all of these elections. We collected data about organizing
roll call votes at the state level. We have some other things in which we are trying to
characterize state politics more generally here. I have also begun to create and possibly
publish someday, at least online, a summary of all of the elections – who the candidates
where, how many ballots they got on various ballots. This is Kentucky which was a multi-ballot
race. So that is what we are doing. Let me show you some big picture patterns,
so back to the flow chart. Of the 752 elections, 510 went the boring way, and 231 went the
fun way with some sort of joint ballot. 19 ended up being deadlocked during this period.
There are 11 elections that we were unable to find any information about. I spent time
in places like Frankfurt, Kentucky going everywhere. Wyoming has never published the journal of
its legislative proceedings of its first legislature when it organizes the state. So there is some
actual fugitive material out there. This shows you the distribution of the actual
number of ballots whenever we get on the side of the flow chart – how many ballots if you
got into the joint balloting side. You see this big spike here. That’s one ballot. And
there that is what you might imagine – split party control of the state legislature so
that a different person wins in each legislature. It goes into the joint ballot, so that whichever
party has a majority wins. That’s the easy case.
There are a few other cases – two, three, four ballots where it takes a couple of days
to work things out. Imagine, Tuesday first vote, Wednesday second vote, Thursday and
Friday of that first week. Sometimes, you go away for the weekend and you come back
– You resolve a few over the weekend, some would drop out or something like that. Then
we have the fun ones where the weekend didn’t help and now you are in the world where you
just trying to figure out – really intractable impasse.
Here is just a note. I’ve annotated the cases here, 7 cases in which there were more than
100 ballots. And when you get here, and these are the ones that people tell stories about
to this day. Five of these end up being deadlocks. Two of these are Delaware. This is the two-year
deadlock, ’99 and ’01. But we did get two elections here. John Logan – there was an
opening and basically a deal between the two parties, then John Palmer in Illinois as well.
But those are the ones where you get the stabbings, bribery. And this is an aside. For the Senate
and you may know that Gerald Gamm and Steve Smith have been writing about the history
of Senate leadership and they note that the creation of the steering committee arose because
of a series of deadlocked elections in 1880, wanted to send delegations of senators out
of the state capitols and try to move some of these things along. These elections were
not only in the news, but Washington would actually descend upon the state capitols to
try to move them along. The darker the state, the more joint ballots
you’ve got. In the West, you’ve got a lot of joint ballots and a lot of fracturing of
the parties in the West. In the Appalachian-Highlands regions, you also see a fracturing of the
parties. But in the West, and they are mostly Senate elections, it took a while to resolve.
This is where all of the deadlocked elections were. And you can see that by-and-large in
the West, in the mountain region, and Appalachia, then Delaware as the problem child. When do
these multi-ballot races come up? We do a bunch of regressions – the sort of stuff that
Richard likes to jump over. Not surprisingly, you’re likely to get a multi-ballot election
whenever the party margins are close. When there is a pivotal third party and the elections
become a three-body problem which we know in physics is impossible to solve and in politics
is frequently impossible to solve. Divided majority parties is another. When a majority
gets really large, the party starts to fracture. Or like the case of the South where you have
different regional strengths. So let me tell you what the take-away is.
I’ll talk about five more minutes and we’ll go to questions. I could talks about this
forever. So this is kind of our take-away – when we kind of drill down into the stories
then look at the big picture. One is that the rise of the parties and the nationalization
of politics in the1820s killed the pure version of the Founders’ Constitution with respect
to Senate elections. By which I mean that by the time that we come on the scene in 1871,
any notion that senators were agents of the state legislature or agents of the state governments
was gone and they become agents of the state parties. Which is true of everything. Presidents
become agents of the state parties. From Rothman and other works, we know that sometimes the
parties were actually the agents of the senators. So sometimes is worked the other way. But
the important thing was that it was the party that was important. Sometimes it was one faction
or the other, but it was the party that was important, not the state government. At the
same time, and this relates to what I said about Riker before, there was not effectively
a popular election for senators. Riker was wrong empirically. Parties were strong and
elections were partisan and insider affairs, so back to what the Progressives really didn’t
like. They didn’t like parties. They didn’t like the opaqueness of the process.
Another word about Illinois, just to keep hammering on Bill Riker. In the book, we look
specifically at Illinois and between the Lincoln-Douglas debate and 1890s, there is no other case in
which the two parties nominate candidate ahead of their conventions. In 1890, John Palmer
goes to the convention and asks to be nominated, but he views that as part of his strategy
to get elected. So this is not what the parties are really doing in the nomination thing.
The general process ended up being this. Rather than having a sort of popular – Riker’s model
which sort of persists in the mind of people was that the parties nominated senate candidates.
There was a state election. If the senator was up for re-election, the state election
would be a shadow election for senators, then the outcome in the state legislature would
determine who the senator was. In fact, what we discover was that during the so-called
popular canvass there may be several people who want the nomination, but there is no one
person on each side of the aisle. There might be pretenders like the primary season now.
Many want the job. They go out on the hustings. They make speeches and they try to make deals.
But during this period as well, the norms of the country are that you don’t really want
to appear to be too eager for the position either. Between the two week period between
the legislature convening and the actually election of the senator, there would be period
of smoke-filled rooms. And if you look at all of these newspapers, you will see some
reporter at the train station watching who was getting off the train, which hotel they
are going to, where the various camps had set up campaign headquarters. And for the
next two weeks there would be the quiet canvass in the smoke-filled rooms. At the end of that
and about the night before the election in the legislature, there would be a caucus where
everything would be worked out to resolve differences. So really the candidates came
from inside the caucuses. The actual results would depend on the particularities.
If the incumbent was running for re-election, usually there was not competition and it was
basically just a confirmation. If the parties were factionalized that would determine the
details. Sometimes there wouldn’t even be caucuses if the parties just couldn’t decide.
And there are reasons they just might not want to have a caucus.
In the book, we look very carefully at New York. We have very good data about where the
candidates come from, what all of the caucus proceedings were, etc. Basically what this
shows is that the various caucus nominations for different people who were nominated..
For instance, in 1881 Hernon who was the incumbent, although now the Democrats are in the minority,
he gets nominated by the Democrats by acclamation. As the incumbent, he’s going to lose, so what
is to be lost. Thomas Platt, the “Little Boss” (with Roscoe Conkling, the “Big Boss”) get
nominated among a fight among the various little bosses of New York. And he wins, 55
to 26 to 10 to 10 to 5. He wins a bare majority. He then goes into election the next and has
all of the Republicans behind him and is elected because his party has a majority. This was
the year that Platt and Conkling get into the tussle with the administration over the
collectorship of the Port of New York. And they end up resigning in protest, assuming
that they are going to be re-elected to the Senate and they weren’t. This was a case where
there never really ended up being a Republican caucus because there was such division between
the Stalwarts and the other faction. They could never really caucus. McKinley gets assassinated
during all of this. And the next day things begin to get resolved behind closed doors.
So that is a case where you never got the caucus, but the Republicans won. We really have two models of candidate convergence
that come from the data. The caucus as post-election mediator between factions where, for example,
Evarts, which I had mentioned previously, his nomination is made unanimous out of a
tame caucus out of which the Democrats could extract nothing. Evarts had a majority wrapped
up, he came in, he won and that was fine. But we actually have roll call votes in the
caucus. The caucus was public. Sometimes there was the caucus as temporizer by which I mean
the parties would fail to make a nomination in caucus. There would be a series of rolling
minorities wasting a lot of votes. If a party can’t decide on a senator the worst thing
that could happen is that you could box yourself into a position where the minority party could
elect the senator. In a lot of these extended, joint balloting, the scattering vote would
be incredible. Everyone get one vote so that there is no chance that shenanigans can happen
and the minority party could win. The Platt-Conkling case is one of those. Kentucky in 1876 is
an example when they started off with five candidates, including John Stevenson which
was the incumbent on the Democratic side – by the way, error in the Biographical Directory
of the U.S. Congress, which says that Stevenson did not try for re-election. In fact, he did
because he was there right at the beginning but he dropped out eventually. Kentucky had
a rule as did a number of other states – the Republican caucus in the House has this rule
as well – if you have a caucus to nominate and there are multiple candidates and no one
gets a majority, then you just keep balloting, the loser drops off, you ballot again and
if no one has a majority the bottom person drops off until you get down to just two people
and you get a majority. Nobody wanted to go into a caucus because nobody who was going
to win there. So Kentucky kept scattering the vote. The caucus refused to meet. And
eventually enough people dropped out, including Stevenson, and they had spent five hours going
back and forth with just two people on the ballot.
I will end by exegeting the cover. It’s 1911. Uncle Sam is looking down at the mess of states.
Here is Iowa. It took 67 ballots to resolve Iowa in 1911. Colorado ended up deadlocking
at 90 ballots. Here is Montana with 79 ballots. And two guys, probably mining owners, battling
over the 79 ballots. Here is New York, two bears pulling 63 ballots. This is the election
in which Senator Franklin Roosevelt shows up as one of the major players on the reform
side during this reform-Tammany fight. In addition, in that election there were 11 other
cases where there was joint balloting. So 1911 was a particularly interesting year from
my perspective. I’ll just so that we’ll have time for questions.
But the bottom line – Thing number one. There is a lot of interesting stuff here. We got
enough for a book, but with respect to data there is a lot more that we would like to
do. Just getting the data out and the accounts of these battles – Theme number two and this
reflects on the movement to repeal the 17th Amendment. It would be instructive to understand
the electoral process in the period that we are studying because I think in many ways
the raw material of American politics now is very similar to the raw material of American
politics then in the sense that parties are very strong. While there is a strong believe,
sincere ideologically, in the power of the states, the parties really are the motive
forces in these elections. The positive, non-political way to weasel out of this is by saying that
if you are interested in giving more power to the states, it is not clear that putting
the election of senators in the hands of state legislatures is the way to do that. That would
just give more power to state parties. And as troubling as national parties may be, we
need to consider state parties. Thank you. Question: At that time, did the Congress not
start sitting until March? Stewart: That is correct. That did not change
until, I don’t recall the exact year, until the 1930s.
Question: Especially with regard to your last point, I am wondering if comparative research
is being done on, say the German system, where the upper house is also a representative of
the states, but it is still a peer-agency relationship because the person sitting in
the upper house does not have an individual vote, but only a vote of the state as instructed
by the state which is in fact, the party ruling the state legislature.
Stewart: I always love it when there is someone who knows the non-American case that challenges
the prevailing thought. Answer number one. No there is not a lot of research going on.
And number two, you just told me why there should be that sort of research. What I would
guess that there is constantly – I would wonder and the question I would have does come from
Ricker, is whether the instructed aspect is a constitutional feature or is the norm in
Germany. What Riker notes is that there had been a norm concerning instructions to senators
early on but it wasn’t written into the Constitution. And further, and I lecture on this to undergraduates
all the time, at the end of the day, the difference between the Constitution and the Articles
of Confederation made it very clear that the Senators were – or could be individuals agents
who followed their constitutional letter. I think that is probably the thing here. And
this is something that is true, probably around 1840, Senators had stopped listening to instructions
and stopped resigning in principle. Senators started saying with respect to instructions:
You don’t like it. Get a majority. You and which majority?
Question: I have a couple of questions. One of them follows from what you just said concerning
the extent to which there really was a true agent-principle relationship. In the classical
principle-agent relationship, the principle can recall the agent any time and for any
reason. And my second question is about the source of the requirement for a majority vote
in the state legislature and whether in a joint session, each member of the state legislature
however many constituents that state legislature had, had the same vote.
Stewart: With respect to the principle-agent relationship, you are exactly right. In a
pure principle-agent relationship, the principle has some sort of sanction like recall or firing
or something like that. And that is an important change in the Constitution from the Articles.
Under the Articles there was recall. Under the Constitution there is not and that is
really an important thing. Except that the norms even without recall, the senators seemed
to take very seriously what the legislature said until around the 1840s. On the source
of majority vote in state legislatures, that ended up being a requirement of this law of
1866. In fact, one of the reasons it was felt this law was necessary was because the states
were beginning to play fast and loose with quorum requirements and things like that.
And there were also questions about whether you had to have a majority in each chamber.
So the law clarified a number of things. In terms of the voting weights – They were one-legislator-one vote, so it wasn’t weighted by the size of your constituency. We know that there was
great variability in the ratio of the lower and upper chambers together, and I have not
done that calculation, but it was certainly really variable. Not surprisingly, by the
way, when there was a bicameral difference, most of the time the winner in the House would
end up winning. But it turns out in about a quarter of the cases the winner in the Senate
won. After all, the important question was not how many votes you got in the chambers
but what was the size of the majorities in the two chambers. And you can kind of come
up with toy examples where you have actually a bigger majority in the Senate than in the
House, although the Senate is the smaller body. And that happened about a quarter of
the time. Question: I had a question and a comment about
the 19 deadlocks. What was the time frame on that? Was that up until the 17th Amendment?
Stewart: Yes. Question: Okay. And the comment was about
your cartoon on the cover. Richard may have mentioned to you that there is going to be
an exhibit here at the National Archives on amending the Constitution and that cartoon
from Berryman will be in the exhibit. Stewart: Excellent! I am glad to know that
I am on the cutting edge of constitutional cartooning. And another thing about the deadlocks.
They came on up to the ratification of the amendment, and it was pretty clear that the
deadlocks was one of the things that was really objectionable to a lot of people – once states
started losing representation. Majorities in the Senate actually took a sort of self-denying
position. The precedent was set very early, maybe even before the Civil War, that if a
state was in deadlock and so there was a vacancy, the Senate would not accept someone appointed
by the governor. Even in those cases where the governor was of the same party as a majority
in the Senate. We know from the work of one of my other collaborators, Jeff Jenkins and
others, that during this period almost every disputed election case was settle on the side
of the party that had the majority in the chamber. But this was an exception. So the
Senate pushed back to the states and said the state legislature has to decide, you cannot
rely on the governor to be the tie-breaker. Question: I was just curious as to what led
you to choose 1871 as your starting point. Stewart: Yes. 1871. And thank you for asking
that because it is something that I wanted to mention at
the very beginning. One answer is that having the law of 1866 is very helpful because we
now have a stereotyped process and now we can start slotting things in. But once we
got into the project, I began to realize that it would have been really nice to go back
to the preceding elections, because it was those preceding elections, going back at least
into the 1850s, gave rise to the idea that the Senate really needed to regulate the selections
more thoroughly. I was giving a hard time to the Progressive Historians and it may very
well be that – a lot of our account of the 1850s are rather fact-challenged. The other
thing which is also really the coward’s way out – and many of you who do original source
work in this area – that there is a real sea-change during the 1860s in terms of the record retention,
printing rather than hand-writing, reporting in a timely way what is happening. The addition
of the telegraph and its spread also helps to convey a lot of information quickly. So
that in the 1860s you can really tell that things really get much better. By around 1870,
you can pretty much get a full record of every election, where my sense is that during the
1860s, problems of the Rebellion aside, it would be difficult in many places to get a
full record of a senate election. I should finally mention, though, that it is not as
if it were a nice, clean break. One of the unfortunate aspects of the project is that
the state of Alabama – The very first election in Alabama involved two competing state legislature,
known as the Capitol Legislature and the Court House Legislature. And the Court House Legislature,
the Republican members, decamped to the federal court house and the Democrats stayed in the
state legislature. And they both elected senators. The journals are not always that good. It
is not clear which journal you are reading. And since Alabama is alphabetically number
one. And this is the election of 1870, the first election in our series, I just concluded
we cannot start with Alabama because I have still not figured out Alabama in 1870. The
very first one that would appear in any compilation – it’s a mess, a royal mess.
Questions: I would just add to what I thought was a most interesting and informative about
your book is the examination of the motivations and justifications of the reform movement
to move from indirect to direct election was how it was going to take the power of money
and corrupt individuals and party politics out of the system and the sense of irony that
it comes back in even stronger form. So the unintended consequences of reform.
Stewart: Those of us who study money and politics know that money will find its way. All money
runs downhill. I should also say very quickly that political scientists have tried very
hard to find the effects of going to direct elections and it is really hard to find policy
effect from going to direct elections. Politics continues more or less as before. One exception.
Franklin Roosevelt probably would not have had the Democratic Senate when he became president
if there had still been indirect election. And the Do-nothing 80th Congress, probably
would have been preceded by the Do-nothing 79th Congress. So there were times when we
would have had Democratic or Republicans Senate that we did not have. But otherwise it is
sort of turtles all the way down. McCulley: Well, Charles on that rather upbeat
note we will conclude the discussion and the terrific talk. Much to think about when we
hear about repealing the 17th Amendment and we thank you for the incredible amount of
research that went into this. It is mind-boggling to think of what you guys have accomplished.
Thank you very much. Stewart: Thank you.

Leave a Reply

Your email address will not be published. Required fields are marked *