The Federal Overseas Voting Law (UOCAVA) Is Unconstitutional

I live in East Lansing, Michigan, and I vote in federal and state elections accordingly. If I were traveling, working, or even living overseas temporarily, I would still vote in East Lansing, Michigan, as an absentee. Being away for a bit, without giving up my permanent Michigan residence, would not deprive me of my vote.

But say that I move overseas, with no intention of ever returning to the U.S. I sell my house in East Lansing. I burn my Michigan driver’s license. I forswear ever drinking Vernors again. I have no contact with anyone back home, and I don’t pay any taxes there. I cannot vote in state elections, because I’m not a Michigander anymore. But under the federal Uniformed and Overseas Citizens Absentee Voting Act (UOCAVA), I can still vote in federal elections, for the rest of my life. Despite my utter lack of ties to my old home, UOCAVA requires Michigan to let me vote as though I still lived in the Eighth District of Michigan. I can vote for U.S. representative, U.S. senator, and for president as though I were a Michigan resident.

Not only is that weird, it’s unconstitutional.

Article I, § 2, cl. 1, one of the very first provisions in the Constitution, provides that the people who elect a state’s representatives in Congress “shall have the Qualifications requisite for Electors of the most numerous Branch of the State Legislature.” In other words, you aren’t allowed to vote for the U.S. House unless you are allowed to vote for the state House. The Seventeenth Amendment does the same for senators: you aren’t allowed to vote for the U.S. Senate unless you are allowed to vote for the state House.

So right away, there’s a problem. Even assuming that Congress can force states to let permanent expatriates vote in their old homes, it seems unconstitutional that Congress does so for federal elections and not state legislative elections. The two electorates are supposed to be congruent, and Congress has violated that requirement.

Presidential elections are more complicated, but the case for UOCAVA is no stronger. In contrast to the electorate for House and Senate, the electorate for president is not defined in the Constitution. This does not give Congress authority to expand the franchise, though. Quite the opposite, it gives the states more leeway to define voter qualifications, or indeed to not have popular voting for president at all. Of course, every state currently uses popular voting, but they can choose different rules at the margins for things like letting felons vote, or (not to assume my conclusion) defining residency.

That said, several constitutional amendments give Congress some control over who votes. For instance, Congress has the explicit power to enforce the amendments that guarantee voting rights—state and federal—based on race, sex, and age. More complicated, but more relevant to UOCAVA, is the Fourteenth Amendment. The Fourteenth Amendment precludes states from violating broad rights like “due process,” which contains multitudes. Section Five of the amendment empowers Congress to enforce these rights. The Supreme Court must agree, however, that what Congress is addressing is really a Fourteenth Amendment violation. Moreover, the Court must agree that the remedy Congress provides is “congruent and proportional” to the violation.

Defenders of UOCAVA typically claim that UOCAVA is a congruent and proportional response to a violation of the right to travel. I disagree.

The right to travel and reside abroad is fundamental and is well-established in the case law. It came up in Oregon v. Mitchell. The federal law in that case allowed states to impose reasonable pre-election registration deadlines—up to thirty days, but no more than that. The Supreme Court agreed that if a state’s electoral system disenfranchised people who arrived more than thirty days before the election, the state’s law violated the Constitution. Conversely, they said, the federal law barring such state laws was justifiable as a protection of the right to travel. Congress further decreed that when voters moved within thirty days of the election and could not register in their new homes, their old states had to let them vote as though they were still residents. The Court agreed that this was appropriate too.

The distinctions with our case are illuminating. The law in Mitchell lets former residents vote in their old states for thirty days after they leave; UOCAVA lets them vote there forever. The people protected in Mitchell are residents of their new states and would be able to vote there if not for the tight timing. They are voters who get to avoid a temporary hiccup. The people protected in UOCAVA are not residents of any state and would be unable to vote if not for UOCAVA. They are nonvoters who are made into voters. As one court put it, the federal statute in Mitchell “was not intended to abrogate the power of the states to enact bona fide residency requirements.” Clearly, the same cannot be said for UOCAVA, which by its design obliterates bona fide residency requirements.

More directly, the right to travel plays out differently in the two cases. The right to travel or move within the United States—the issue in Mitchell—is much stronger than the right to travel or move abroad, which the Court has allowed to be restricted in many more ways. There is a limited constitutional right to go abroad, in other words, and the Constitution lets the government impose costs on the exercise of that right. Benefits that aren’t tied to residence in a certain place (social security, for instance) are one thing, but things tied to residence are different. And voting is tied to residence in a state, by tradition, by law, and by the Constitution. Representatives and senators, for whom UOCAVA lets expatriates vote, are supposed to represent particular places, or at least people in particular places. So too are the members of the electoral college who, state by state, elect the president. Permanent expatriates who vote in their old states are not counted as residents of the state for any other purpose, including state taxes, jury duty, or voting in state elections. They are not counted in the determination of how many representatives and electoral votes the state gets.

Even if it would violate permanent expatriates’ right to travel and live abroad for their states to bar them from voting, UOCAVA is probably not a “congruent and proportional” response to that violation. Temporary expatriates with a permanent address back home might have a claim that their constitutional right to travel is violated if they are disenfranchised. It is hard to argue that this justifies giving permanent expatriates the vote as well.

A further congruence problem is that UOCAVA gives permanent expatriates the wrong kind of vote. Even to the extent that these people have a continuing interest in being represented in the federal government, it is hard to see how that interest translates into voting for a representative and senators from a place where they no longer live. As already noted, members of Congress represent places as they represent people. The vote that UOCAVA provides permanent expatriates is mismatched if the point is to have their interest as expatriates represented. (If that isn’t the point—if we are assuming instead that expatriates should be represented if they have a continuing interest in the places they used to live—then the point would be moot, because it would be fair to use some sort of residency requirement to gauge somebody’s continuing interest in the community.)

UOCAVA’s defenders could note that presidents, at least, represent the whole country rather than individual constituencies. Moreover, the interests of expatriates are protected most directly by the watchful eyes of the president’s diplomatic corps and embassies, not by anything Congress does. But we do not have a simple popular vote of all citizens to elect the president. Presidential elections are carried out on a state-by-state basis, and votes for president are cast by voters as citizens of various states, and then by electoral-college members who represent each state. Until and unless the electoral college is abolished, presidential votes under UOCAVA are just as problematically un-geographical as congressional votes under UOCAVA are.

One final absurdity in UOCAVA’s design provides the most damning evidence that there are no constitutional rights being protected here. There are plenty of U.S. citizens who don’t get to vote. Residents of Washington, D.C. participate in presidential elections, but only because of a constitutional amendment. Resident citizens of Puerto Rico, Guam, and other possessions cannot vote for president or Congress at all, despite being U.S. citizens.

This sets up a truly odd incongruity. If an American citizen moves to North Korea with no intention of ever coming back, UOCAVA gives him the right to keep voting in federal elections. If he instead moves to Puerto Rico, he cannot vote at all. So much for the right to travel.

The courts have upheld UOCAVA’s lack of coverage for people who move to Puerto Rico, which deflates the argument that UOCAVA is needed to redress a constitutional violation. If you can lose your “right” to vote for president or Congress by moving from Michigan to Puerto Rico, then your right to travel and your right to vote must not amount to much. If Michigan can take away your vote in that situation without violating the Constitution, surely it can take away your vote if you move to North Korea without running afoul of the Constitution. If there is no constitutional violation, then Congress has no power to barge in and overrule Michigan in the name of defending your purported Fourteenth Amendment rights. The right to travel does not justify UOCAVA.

Think that this stuff doesn’t matter? Consider the 2000 presidential election in Florida. While pregnant chad dominated the news coverage, the less prominent issue of overseas ballots decided the election. Twice. After the initial, uncontroversial recounts, George W. Bush had a slim lead of three hundred votes. He owed his lead to overseas voting; tens of thousands of overseas ballots had arrived early enough to be counted on Election Day, and Bush’s advantage among them was more than three hundred votes. Florida law allowed ten more days for overseas ballots to arrive, so there were still thousands of votes to count. Bush added a relatively whopping 739 net votes to his lead from these late-arriving overseas votes. After all of this (and all of the other recount convolutions), Bush’s victory was declared by a margin of 537 votes. The overseas votes had saved Bush again.

You may also like...

20 Responses

  1. Adam says:

    It may be unconstitutional, but I’m more impressed that you used “Michigander” in your blog posting. I think this is the first time I’ve seen that term in a blog posting. Woohoo!

  2. A.J. Sutter says:

    For what it’s worth: I moved to Japan about 2 years ago from San Jose, CA, and don’t have any present intention to return to US. However, when I wanted to register for the 2008 election, I had to do so through the Santa Clara County (CA) board of elections. (I declared, by the way, my lack of intention to return to US.) The absentee ballot I was sent included measures not only for state offices such as state legistature and executive positions, but even ballot measures and local races pertinent to the San Jose neighborhood where my last address was, e.g. for the Evergreen local school board.

    I haven’t checked the UOCAVA, but in my California experience there seems to have been de facto compliance with Article I, § 2, cl. 1 of the US Constitution. If de facto compliance is indeed the norm, that might explain why (I presume, since you don’t mention and I haven’t checked) there haven’t yet been challenges to the law. Is Michigan different in fact?

    I admit it felt a bit weird to be asked to vote on micro-local issues; I generally abstained on those. Perhaps the reason for these being on the absentee ballot, though, include:

    @ US and the state (a) have an interest in hoping I will change my mind someday about being a permanent expat, and (b) they figure I most likely would return to where I came from, so for both of these reasons they want to keep me engaged with local issues; and/or

    @ it might be problematic to have separate absentee ballots for people who say they want to return and people who say they don’t. (Consider, too, that some people in each group probably will change their minds in the future.)

  3. Brian Kalt says:


    If every state chose to do what you describe California as doing, there would be no problem. Indeed, states are perfectly free to choose to define residency as including people in your situation. The problem is not expats voting, the problem is Congress forcing states to let expats vote.

    UOCAVA provides, incidentally, for a simple federal ballot for people to use. In states where the expats can only vote in federal elections, or where it is not feasible to get the regular ballot (which includes state races) out to the expat, the simple federal ballot is used.

    In your example, California still could, saying that it lets expats vote in state and local elections because Congress forced inclusion in federal elections, and the state didn’t want to then violate the symmetry requirement. I theorize in more detail in my book about why states have not challenged this law. The bottom lines are: (1) this is also the law that provides for overseas military voting, and it would be politically tricky to try to extricate military voters from the challenged permanent expats; (2) I’m guessing that states don’t mind having their proper roles usurped–at least not at the margins.

    Your last point, about distinguishing between people who intend to return and those who don’t, is at the root of the problem. Predecessor laws to UOCAVA covered only people who would otherwise be qualified to vote, but just happened to be abroad for the moment. Rather than get into the issues of proving who was going to return and who wasn’t (which in theory would only require maintaining a permanent address, but in practice could get harder to prove), Congress just said that everyone should be allowed to vote.

    Personally, I don’t think that’s a good enough reason, but that’s just me.

    Anyway, thanks so much for your data point–as I write this up, it helps to have individual perspectives like that.

  4. A.W. says:


    i concurr completely. When i was in law school i refused to vote for anything besides president on moral grounds. i was a texan going to school with connecticut with the intention to live in the DC area after school. It didn’t seem right to have any say whatsover in Texan issues, because i didn’t intend to return, or CT issues, because i didn’t intend to live with the consequences of my decision for any appreciable time. I have no respect for people who leave this country for good and then think they somehow retain the right to meddle in our affairs.

    That’s a policy argument, rather than a pure legal one, but i wanted to get that off my chest.

  5. Jason says:

    Professor Kalt:

    Once again you find some of the most interesting things to write about. I’m quite the fan.


  6. jt says:

    UOCAVA is perfectly consistent with Michigan law.

    Michigan allows a U.S. overseas citizen who is 18 years old, not registered to vote anywhere else in the U.S. and who is a spouse or dependent of a Michigan resident to register and vote in Michigan elections even though they have never established Michigan residency.

    Michiganders can relax.

  7. Brian Kalt says:

    jt, my whole point is that Michigan does that only because UOCAVA forces it to.

  8. A.J. Sutter says:

    Brian, could you please expand on the facts a little? It’s not immediately obvious why California has “chosen” to let me vote, but Michigan would be “forced” to let you vote. How are the situations different? Do Michigan absentee ballots not include state and local offices?

  9. John says:

    Doesn’t this ignore the effect of Art. I, Sect. 4, giving Congress broad authority to pass laws trumping requirements Sect. II would otherwise mandate? (See Black’s opinion in Oregon v. Mitchell.)

  10. Brian Kalt says:

    A.J., sorry for the confusion I caused by not reading jt’s comment closely enough. Here’s the scoop. UOCAVA forces states to let permanent expats vote in federal elections. A state, having already been forced to let the permanent expats vote in federal elections, might choose to maximize symmetry and let those same folks vote in state and local elections. But that doesn’t mean that the state wouldn’t rather bar those folks from voting altogether, if they had the option.

    What I missed was that jt was saying that Michigan lets permanent expats vote in state and local elections as well as federal ones. I don’t think that’s right, but I can’t say that I am positive.

    jt, do you have a citation or link that you can direct me to on that one? All I have seen says that only people who reside in Michigan and are temporarily away (and their spouses and dependents) can vote in local elections, not that permanent expats can.

    John, I did ignore that point, because my post had gone on quite long enough, I thought. But your point is a very incisive and important one. My short answer (which I cover in more depth in my book chapter) is that Justice Black’s position has never been adopted by a majority of the Court, and I don’t think it is the best reading of that clause. The clause gives Congress the power to regulate the “manner” of congressional elections, but I take manner to be about procedures, not about defining the electorate. And even if “manner” did include that, it wouldn’t extend to presidential elections, in which the “manner” is left to the states.

  11. Alan Gura says:

    Interesting. I wrote an article about the unconstitutionality of this law for the Texas Review of Law & Politics back in 2001.

  12. Brian Kalt says:

    Alan! How are you? Yes, and a fine article it was. For what it’s worth, I draw heavily upon it in the book chapter on which this post is based.

  13. jt says:

    My earlier post was pasted from the FVAP (federal voting assistance program) web site and was not (blush) based on statutory research. I do not know the provenance of the FVAP statement or whether MI does this by statute or regulation or just does it. If by statute, I think there is no constitutional issue. More boradly, my untutored thinking is that everyone — those whose homes have been foreclosed or hwo are washed away by hurricanes — has domicile in a particular place (voting precinct) until she establishes domicile in another particlar place. If not, I don’t know that UOCAVA necessariy must be read as allowing persons not eligible under state law to vote. It has specific limitations to comply with state law voter qualification requirements in most – but not all– of the definitions, and a court might well read that requirement into the act as a whole — at least in the context of a facial challenge. If not, in real life there probably remains the problem of finding someone who has, in effect, renounced her ties to Michingan and yet continues to vote there and is willing to so testify.

  14. Alan Gura says:

    Doing fine, Brian, thanks. Glad you found the article useful. It seemed to be an obscure issue even at the time. But I don’t see any willingness to litigate it, both parties think they can game the system the way it is, and nobody wants to be “against voting,” although allowing unauthorized voters dilutes the votes of everyone else…

  15. Brian Kalt says:

    jt, I went to the FVAP site, and I can’t deduce from the statement whether they mean that expats can vote in state and local elections or not. I don’t think that they can.

  16. Jens says:

    “And voting is tied to residence in a state, by tradition, by law, and by the Constitution. Representatives and senators, for whom UOCAVA lets expatriates vote, are supposed to represent particular places, or at least people in particular places.”

    The last sentence is a non sequitur for me. Elections are based on a regional system, ok. But that does not mean that the Senate and the House as a whole represent the People. And that means that people who do not fit in the regionalised model have to be put just somewhere. And that is then the region to which the person in question has the closest ties (as further defined by some abstract, manageable criteria like last place of residence).

    I think the view on this issue might be influenced by nationality law: The US have mainly ius solis law (I think the ius sanguinis component requires registration?), so you define the electorate by the connection with the soil. For countries with ius sanguinis nationality law, it’s just “the people”.

  17. Brian Kalt says:

    Jens, let me put it this way. My senators represent the people of the state of Michigan, not the people of the state of Michigan plus the people who used to live in Michigan minus the ones who moved from here to somewhere else where they can vote. Just the people of the state of Michigan. I just think that’s part of the fundamental structure of the Senate that the Constitution establishes, and I don’t think that Congress has the power to change that through ordinary legislation. Michigan does, uncoerced, but not Congress.

  18. Jens says:

    “My senators represent the people of the state of Michigan,”

    And you loose Michigan citizenship if you leave Michigan? So at least state citizenship comes only with the soil, but not with the blood? I mean, the notion of separate “state peoples” stems from the fact that the states entered the union with their sovereignty intact, right?

    It’s just all a bit surprising and interesting to me. I think the constitutions of the German Länder generelly do not provide for a separate state citizenship, and article 33 paragraph 1 of the Basic Law provides: “Every German has the same citizenship rights and duties in every Land.”

    Well, I think I got the crucial points: That the US constitution leaves competences for voting laws for federal elections with the states. The only other example for that I can think of right now is the direct elections of the members of the European Parliament. Interestingly, you can vote in any (one) member country whose nationality you have or (exclusive or) where you have your permanent residence.

    Btw, an expatriate Austrian who has no specific connection to any Austrian municipality can vote for the Vienna magistrate (in voting district 1). At least that’s what a bi-national colleague told me, I haven’t checked.

  19. Brian Kalt says:


    Section One of the Fourteenth Amendment begins “All persons born or naturalized in the United States, and subject to the jurisdiction thereof, are citizens of the United States and of the State wherein they reside.”

    So yes, I think that residence does carry a lot of weight when it comes to state citizenship.

  20. citizenw says:

    Some of the issues raised here also relate to problem of the lack of voting and representation in DC. If one moves from a State to a foreign country, one can still vote and be represented in that State of origin, indefinitely, but if one moves from a State to Washington DC, which is arguably not the one of the States, one can no longer vote and be represented in the State of origin. There exists an ironic incongruity here, which the Courts have refused to recognize.