Your vote is probably taken for granted. It doesn’t have to be that way.

The 2016 presidential campaign made it pretty clear that unless you live in a battleground state, candidates and media outlets will mostly ignore you. There’s a good reason why, and it’s baked into the U.S. Constitution. But there’s also a way to fix it — one that can strengthen our democracy and improve the campaign experience for everyone.

When you mark your ballot, it doesn’t really decide who becomes president. It’s a state-by-state vote that culminates with the Electoral College actually making the decision. As a result, presidential campaigns figure out which states where they’re competitive with their opponent in the polls and focus the bulk of their efforts there (these are the “battleground” or “swing” states).

That’s exactly what happened in 2016:

Two-thirds (273 of 399) of the general-election campaign events in the 2016 presidential race were in just 6 states (Florida, North Carolina, Pennsylvania, Ohio, Virginia, and Michigan). 94% of the 2016 campaign events (375 of the 399) were in 12 states (the 11 states identified…as “battleground” states by Politico and The Hill), and Arizona.

So if you’re a Republican living in California, where 60%+ of voters will vote for the Democrat on the ticket, neither presidential candidate is much interested in campaigning near you — because absent a once-in-a-lifetime type election, the Golden State’s electors are all going for the Democratic candidate. Likewise, Democrats in Texas know the Republican candidate is probably getting that state’s electoral votes no matter what.

Meanwhile, voters in swing states can’t surf the web — much less choose a TV channel — without being overrun by campaign commercials. They’re mailed, phoned, door-belled, and invited to town hall events with the candidates or their surrogates…because their vote is actually sought-after.

This is less than ideal for all concerned. Whole swaths of the country never see the candidates, and a relatively tiny portion gets inundated by them. And it’s all because the path to winning the presidency — win the battleground states to win the Electoral College — is so narrow.

One way to reveal more nuance in the vote is to use not just two colors, red and blue, but to use red, blue, and shades of purple in between to indicate percentages of votes. Here is what the normal map looks like if you do this. (Source, courtesy of Mark Newman, Department of Physics and Center for the Study of Complex Systems, University of Michigan)

As this version of the map (scaled by population) makes clear, large portions of the country are quite evenly divided, appearing in various shades of purple, although a number of strongly Democratic or Republican areas are visible too. (Source, courtesy of Mark Newman, Department of Physics and Center for the Study of Complex Systems, University of Michigan)

Just over 120 million votes were cast for president this year, so under a national popular vote, the winning candidate would need slightly more than 60 million votes to win. But no state has more than 10% of the nation’s total registered voters (according to 2014 U.S. Census Bureau data) — and most states have a sizable numbers of voters on both sides of the political spectrum: in 2016, the losing presidential candidate still got more than 33% of the vote in 39 states.

Put the Census data and 2016 election results together, and it’s apparent that reaching a 60 million majority vote threshold under a national popular vote would have required both Clinton and Trump to campaign in some rather unconventional places by today’s standards.

For Trump, the states most rich in Republican voters are: Texas, California, Florida, New York, Pennsylvania, Ohio, Michigan, North Carolina and Illinois. (Notice two “blue” heavy-hitters there, California and New York, as well as several contemporary battleground/swing states.) But those 9 only get him the first 30 million votes he needs.

The next 30 million votes he needs are in 22 more states: Georgia, Tennessee, Missouri, Virginia, Indiana, New Jersey, Alabama, Wisconsin, Kentucky, Louisiana, Arizona, South Carolina, Minnesota, Washington, Colorado, Massachusetts, Oklahoma, Maryland, Mississippi, Oregon, Iowa and Arkansas. A lot of those are majority Republican states already. But New Jersey, Minnesota, Washington, Colorado, Massachusetts, and Oregon are also there along with several swing states.

Clinton’s path to her first 30 million votes is shorter: California, New York, Texas, Florida, Illinois, Pennsylvania and Ohio. But note that seven of those states are also on Trump’s list (more on that below) — and that’s still only halfway to the total she needs to win.

Her next 30 million votes would most likely be found in these 20 states: Michigan, North Carolina, New Jersey, Massachusetts, Georgia, Virginia, Washington, Maryland, Wisconsin, Minnesota, Colorado, Arizona, Missouri, Indiana, Tennessee, Oregon, South Carolina, Louisiana, Connecticut, Alabama. Yes, some of those are already majority Democratic— but there are a number of swing and red states on her list where she’d also be likely to campaign.

The most common criticism of a national popular vote is that candidates would campaign only in populous states, and ignore all the others. But the math doesn’t favor that campaign strategy. If you’ve been keeping track at home, Trump would have 31 states on his list; Clinton, 27. And they’d have 26 states in common — only Kentucky, Oklahoma and Mississippi are on Trump’s list alone, as is Connecticut on Clinton’s.

Under that scenario, put yourself in the shoes of a presidential campaign manager for either candidate. Wouldn’t you be wondering: is our opponent poaching voters in states we haven’t visited yet? Is his (or her!) case for election resonating with voters in a state we haven’t visited yet? How do we configure our ad buys? Where should we set up another campaign office?

Under a national popular vote, presidential candidates would have good reasons to go get votes in many states that are entirely ignored now. And voters would have better reasons to turn out for their preferred candidate — not just for the election, but also as donors and volunteers during the run-up to it.

A national popular vote would result in a more dynamic and fluid campaign: candidates would have to listen, and make their case, to many more voters in many more states. A candidate wouldn’t need to win a majority in a targeted state to succeed— just win more voters than projected from their rival.

Does eliminating the Electoral College require amending the Constitution? That’s one path. But it’s really not necessary, because each state decides how to apportion its electors. So if any group of states with 270 collective Electoral College directs those electors to vote for the winner of the popular vote, that effectively institute a national popular vote for president.

That’s the premise behind the National Popular Vote project, which would guarantee the Presidency to the candidate who receives the most popular votes in all 50 states and the District of Columbia.

The legislation to do so, called the National Popular Vote Interstate Compact, has already been enacted by 11 jurisdictions possessing 165 electoral votes — 61% of the necessary 270 electoral votes. That includes four small jurisdictions (RI, VT, HI, DC), three medium- size states (MD, MA, WA), and four big states (NJ, IL, NY, CA). All told, it has passed in a total of 33 legislative chambers in 22 states — most recently by a bipartisan 40–16 vote in the Arizona House, a 28–18 vote in the Oklahoma Senate, a 57–4 vote in New York Senate, and a 37–21 vote in Oregon House.

We could have a national popular vote for president by 2020. To learn more — and get tools to lobby your state legislature to enact National Popular Vote legislation— visit http://www.nationalpopularvote.com/.