Senator Russ Feingold recently proposed a Constitutional Amendment which would mandate special elections for Senate vacancies similar to those required for those in the House of Representatives.  Below are resources produced by FairVote analysts that explain the problems with gubernatorial appointments for Senate seats and the benefits of using special elections for Senate vacancies.

David Segal's Sunday New York Times Commentary

David Segal's Congressional Testimony

David Moon's op-ed: Fumbling toward democracy



Senators Russ Feingold, John McCain, Mark Begich, and Richard Durbin proposed SJR 7, to amend the Constitution in order to require all U.S. Senators to be elected, including when seats become vacant.  This information sheet explains why this legislation is necessary and addresses concerns of feasibility and logistics.

 
What SJ 7 does

SJ 7 closes a loophole in the 17th amendment of the U.S. Constitution that established popular election of U.S. Senators. That amendment allowed states to fill vacancies for U.S. Senate seats without elections. States already had procedures to fill vacancies without elections and governors immediately recognized the value to them of being able to appoint vacancies themselves. As a result, nearly all states left governors with the power to fill vacancies. In the nine decades since passage of the 17th amendment, governors have appointed nearly a quarter of all U.S Senators.

Specifically, SJ 7 uses language similar to the constitutional language requiring states to hold elections for U.S. House vacancies.  The amendment states that:

Section 1. No person shall be a Senator from a State unless such person has been elected by the people thereof. When vacancies happen in the representation of any State in the Senate, the executive authority of such State shall issue writs of election to fill such vacancies.

Section 2. This amendment shall not be so construed as to affect the election or term of any Senator chosen before it becomes valid as a part of the Constitution.

The status quo is unacceptable

Former Illinois governor Rod Blagojevich  taught us in 2008 that anything can happen in the process of filling a vacant US Senate seat.  While Illinois circa 2009 provides an especially stark example, it is always problematic to give a single person so much power.  Since 1913, nearly a quarter of U.S. Senators have been chosen by gubernatorial appointment – more than 180 in all. Too many of them have entered office under clouds of suspicion, undercutting their ability to successfully advocate on behalf of their constituents.

There is no reason to believe that the wants of a governor will consistently align with the will of the people he or she governs:  The appointment of senators by governors is often characterized by back-room wheeling and dealing, underlined by any of a variety of nefarious motives: consolidation of power, friendship, kinship, a drive for personal enrichment, or otherwise.  (Note the reasonable speculation that one purpose of Blagojevich's pick of Burris was to influence potential jury pools.)  Voters choose governors based on consideration of local issues, rather than federal issues -- as such, a number of states regularly have governors and senators of differing political affiliations.  There is simply no reason to believe that governors are better equipped than are the people, in determining states' representation in the U.S. Senate.

An amendment is necessary

State legislation to establish vacancy elections is useful, but it is unlikely to yield widespread adoption of this important reform.  Governors and legislatures in states where a single party dominates the government will strive to retain governors' power to pick U.S. senators of their choosing, particularly if they believe governors in states controlled by the other major party are not planning to cede such authority. States with divided control are unlikely to pass legislation unless a veto-proof majority of one party controls the legislature, while the governor is of different party affiliation.  (In all scenarios, governors can be expected to fight tooth-and-nail to maintain their power.)

Elections won’t require that seats be left vacant for an unduly long period of time
        
Special elections have governed U.S. House vacancies since the establishment of Congress, and a handful of states have made regular use of special elections for U.S. Senate vacancies, yielding no major problems that we have been able to uncover.  In addition, our society has regularly accepted senators' missing large amounts of time when ill.

One or two extra months spent preparing for an election is well worth the wait, and this interval could be truncated through the use of instant runoff voting (IRV).  As practiced in some state legislative vacancy elections, states could explore allowing a senator to announce his or her resignation prospectively, and hold special elections prior to the effective date of said resignation.  (This would have been useful in November of 2008, when Senators Obama and Biden were elected to higher office.)

Further, quick appointments are no assurance of due representation: There is reason to believe that appointed senators typically serve with less clout than do their elected colleagues; An appointed senator is less accountable to his or her constituents, and is with diminished obligation to act in their interests.

Special elections will not yield a unique advantage to candidates with more money or better name-recognition

Candidates with broad name-recognition and/or high fundraising capacity already have an advantage in electoral politics; history shows that such candidates have no special advantage in vacancy elections as compared to regularly scheduled elections.  Even where these advantages are present, the election of a candidate who benefits from them is preferable to a choice made with little or no public input -- and high name-recognition should be a prerequisite for holding office, insofar as it is indicative of voters' familiarity with a given candidate.

Concerns about crowded fields or low turnout in special elections can be addressed

Special elections won’t necessarily draw larger fields than regularly scheduled elections for open seats, but any such concerns can be mitigated.  For example, for U.S. House vacancy elections, New York allows parties to nominate one candidate through its own internal processes. Louisiana and Mississippi use runoff elections, while Vermont’s special election bill proposes use of IRV.
Special elections may have lower turnout than regularly scheduled elections, but even low-turnout elections are more democratic than no elections at all.  States can explore using innovative methods like IRV, vote-by-mail, or weekend voting to encourage greater participation.

Special elections need not be too costly for states

A special election might cost a few dollars per voter, but this should be considered money well-spent.  An honest, effective U.S. Senator can earn or save his or her state hundreds of millions of dollars; conversely, governance run amok can be quite costly.  There is a qualitative cost – in terms of increased cynicism and decreased likelihood of future participation in democracy – to allowing for appointed, unaccountable governance. It would be speculative, but not unreasonable, to contend that such decreased faith in the legitimacy of government has other, more quantifiable, ramifications – such as decreased propensities to pay taxes, obey laws and volunteer for military service.

In recognition of the federal government's interest in how senators are elected, it would be reasonable for Congress to set aside a pool of money to fund such special elections in whole or in part.