On Sept. 13, the Boston Globe published the results of a Globe/Colby College poll that showed that the presidential race in Maine had narrowed. A lot. Back at the beginning of August, one poll showed Hillary Clinton with a ten percent lead over Donald Trump in Maine, and as recently as Sept. 5, one poll showed the Clinton campaign with a nine percent lead. Now, that lead has apparently shrunk to three percent, which is within the margin of error. And while initially I thought, “well, one poll…,” Amy Walter, at the respected, non-partisan Cook Political Report, agrees that Maine is now in play in the presidential election. Beacon’s own tracking poll, the UNH/Portland Press Herald poll and a recent Washington Post/Survey Monkey poll all now tell a similar story.

Part of this story has to do with the divide between southern and coastal Maine and northern and western Maine; Trump is currently leading Clinton by 10 percentage points in the Second Congressional District while she leads him by eighteen points in the First CD. That’s an important story, but I’m going to have to focus on that in another column, because today, I want to address another part of the data from these new polls: the third- and fourth-party vote, and the so-called independent vote.

The Globe/Colby poll shows Libertarian candidate Gary Johnson with 9 percent of the vote, and Green Jill Stein with 5 percent; in the Post poll, the numbers are even more striking — Johnson has 15 percent there, Stein 8. (Ordinarily, the Libertarian candidate would draw votes from a Republican candidate — Johnson’s platform represents Americans who look fondly at the 18th, or even 18th, century, calling for the abolition of Social Security, environmental regulation, child labor laws, Medicare, and public schools — but, as you may have noticed, this election year isn’t ordinary, and he seems to be drawing as many voters from Clinton as from Trump, if not more.)

Maine voters who want to vote for Clinton because they love Clinton or hate Trump, or who want to vote for Trump because they love Trump or hate Clinton, aren’t really who I’m talking to here. If your vote choice is premised on what the top of the ticket offers (experience and a life of public service versus the Triumph of the Will, essentially), then we’re good here. It’s the 12-23 percent of the Maine electorate looking to vote for Johnson or Stein that I’d like to address.

The personality of individual presidential candidates matter, of course; but they matter in particular contexts. Back in the last century, after the Democratic Party exploded on live television during its national convention in Chicago, Democrats adopted a system of primary elections and grass roots nominating caucuses. The idea was to empower individual members of the party to play a role in choosing their party’s nominee for president, rather than leaving the process to party elites. The experiment in broad-based democratic participation worked, kind of. That is, the reforms had effect on the nominating process, but that effect brought on many unintended consequences. To win either party’s nomination now, candidates must hire their own campaign staff, communications people, and pollsters, and they must find a way to mobilize majorities of their party’s electorate in state after state.

See, we talk about our parties like they are organic, cohesive beings, with consistent personalities and character traits, when, in fact, they are complex patchworks of coalitions. In winner-take-all elections systems like ours, bundling as many votes as possible into the same bucket just makes mathematical sense, no matter how difficult it is to manage the relationships among all the groups that it takes to build a majority. Winning nomination, therefore, requires pleasing many groups within the party, spread out over a vast amount of territory. This is how Clinton was able to win the Democratic nomination despite what often seemed, to those comparing rallies, like Sanders had the majority of energized crowds: she won over a majority of all the different factions within the party, while he won the support of a few, really big, factions.

Now, if Clinton wins office, she is going to need to work with those people to successfully do her job. Members of some groups might be asked to join her cabinet; policies dear to other groups will be emphasized in speeches, and, ideally, in policy proposals. We won’t just be dealing with a President Clinton; we’ll be dealing with representatives of the NAACP, the Human Rights Campaign, Planned Parenthood, and Everytown for Gun Safety. Clinton has detailed a variety of policy goals, including abolishing in-state college tuition for people from families that earn less that $125, 000 a year, reforming the nation’s immigration laws with an emphasis on family unification and legal paths to citizenship, and passing the Paycheck Fairness Act. These policies may be unlikely, but without the support of members of her broad, complex, national party and the interest groups which make it up, the passage of any of these plans (or even of routine budgets) will be all but impossible.

Trump has lost the support of important opinion-makers in the GOP, including both Presidents Bush and our senior senator, Susan Collins, but clearly he was able to pull together a winning coalition in the primaries. His support of racial profiling, stop-and-frisk, and a wall on the U.S./Mexico border has won him the support of the Fraternal Order of Police, the head of the American Nazi Party, and a host of “white nationalist” organizations and advocates. But despite the fact that these groups show little experience dealing with legislative battles, and despite the disorder prevalent within the GOP caucus of the House of Representatives, Trump would still have national coalitions of elected officials and activists to aid him in passing his agenda.

The Libertarian and Green Parties do not have multiple overlapping national groups offering political support; they do not have a patchwork of grass-roots interest groups and local elected officials; and they are not represented in Congress. If you support, for example, the Johnson/Weld policies on doing nothing regarding global warming because the earth will fall into the sun one day anyway, you will not be able to see a legislative strategy built around those ideas moving through government.

And, more importantly, if you do not support such policies, why would you support them? Americans see themselves as individuals, and thus can be forgiven for seeing their votes as tiny and unimportant pieces of a larger puzzle that really only have legitimacy if they increase the representation of the ideas near and dear to the voters in question. But voting isn’t an individual activity, secret ballot aside. Voting is part of the complex national relationships that make up our system of government. It’s not the place for theatrical gestures in honor of the left- or right-of-center candidate who you like most. Consider the relationships that contribute to passing law, enforcing law, or, on the Supreme Court, interpreting law, and consider which candidate is part of the relationships that are central to your worldview.

This isn’t just about you, me, Clinton, or Trump; it’s about the millions of people whose lives depend on the way people throughout our government do their jobs. Please consider that, and vote accordingly.

Ron Schmidt is an associate professor of Political Science in the Department of History and Political Science at the University of Southern Maine. The preceding originally appeared on mainebeacon.com, a website and podcast created by progressive group the Maine People’s Alliance.

Comments are not available on this story.