The U.S. Senate is at the epicenter of what is shaping up to be an interesting political year. If things hold to form, the Republicans stand a reasonable chance at gaining control of the Senate and keeping control of the House. This would effectively clip President Obama’s already fragile political wings. Let’s take a look at where things stand now and then at some historical trends to keep in mind as we think through what might happen in November.

Our tendency is to read national events into local contests. But every Senate campaign is a local election, even though it involves an entire state. While there are national issues that can creep into, and sometimes even define, a local contest, it is not necessarily the case. Senate elections combine the small-scale of the state with the larger scale of the national political environment, but that combination varies because the local context varies, sometimes dramatically, from state to state.

Think of most elections as a delectable New Orleans gumbo. All gumbos start with a “roux,” or base. On top of the roux, you can add a variety of ingredients–okra, shrimp, duck, celery, garlic, and sausage, just to name a few–but what you toss into the pot is up to the chef and it depends on the materials at hand. For this reason, if you travel through the South, every gumbo you encounter will taste the same but different. Just like elections.

Since the contests are fundamentally local, we have to think of the local dynamics first, just like the gumbo’s base. These are easily knowable and predictable things in some ways. What are the demographics (race, age, social class, and education levels) at play? How competitive have the two parties been in this state or area? Is there an incumbent running or not? All of these things are pretty straightforward, but they vary from state to state. This also means that national issues, if they creep into the contest, will manifest themselves differently based on these factors. To use an obvious example, immigration plays pretty differently in Arizona as compared to Rhode Island.

All that to say, generalizing a bunch of local contests into a national trend is fraught with danger, both practically and methodologically.

Senate Basics

The Democrats currently control the Senate with 53 Democrats, 45 Republicans and 2 independents. The independents caucus with the Democrats, giving them an effective majority of 55-45. To take control of the Senate, Republicans need to pick up six seats–remember, in an evenly split Senate (50-50), the President of the Senate (the V.P. of the United States) casts the tie-breaking vote, so Democrats would maintain majority control unless they lose at least six seats.

There are 36 seats being contested this fall. Of those, 21 are Democrat seats and 15 are Republican. The Republicans have fewer seats to defend, but this only matters if some of the Democrat seats are vulnerable.

Vulnerability

Are there good reasons to think that some of these 36 seats might flip to the other party?

Historically, incumbents do quite well in U.S. Senate elections, generally winning 80% or more of the time. The best chance for seats to flip, then, is when the seat is open and no incumbent is running. There are 7 open seats in this election. Though it is exceedingly unlikely, if the Republicans win all of the open seats, they will take control.

Vulnerability is not limited to open seats, however. Even incumbents lose, but when? Most obviously, scandals can be bad for incumbents. Also, the local political context that surrounds the candidate might shift, making them out of step with their political environment. For instance, if a state has been historically Democrat, let’s say, but is trending toward Republicans in more recent contests, Democrats might become increasingly vulnerable. This is one way in which a six-year term can hamper a Senator. Finally, national political trends can weigh down or boost partisans, often in ways beyond their control. Weak candidates may be bolstered by a tide in their party’s direction, or strong candidates could struggle if the national wind shifts.

Open Seat Breakdown

Of the 7 open seats, 4 were previously held by Democrats (WVa, MI, IA, SD) and 3 by Republicans (GA, NE, OK). When we look at these seats, are there particular reasons to think one or other party might be at a disadvantage?

Of the 4 Democrat open seats, Mitt Romney, the Republican presidential candidate in 2012, actually won two of them (SD, WVa). Of the 3 Republican open seats, President Obama won none of those states in 2012. This suggests at least the possibility that the open seats might tilt toward Republicans, all other things being equal. Georgia could be the one shock for Republicans here. While I think they will still hold onto the seat, the Democrat challenger, Michelle Nunn, has a legendary political name. Her father, Sam Nunn, was a stalwart Democrat Senator in the Peach State and if Nunn can cast herself as a moderate, she might have a chance.

Shifting Local Political Contexts

Of the remaining 29 contests, those where an incumbent is running, there are several incumbents in serious trouble.

In Montana, the “incumbent” is actually a recent appointee, John Walsh, who replaced a retiring Democrat. Montana went for Romney with 55% of the vote in 2012 and the state has been strongly Republican at the presidential level for several election cycles. This looks like a potential Republican win.

In Alaska, Arkansas, and Louisiana we have Democrat incumbents who reside in states that have been trending more Republican during the last several cycles. (Mitt Romney won all three of these states in 2012.) These incumbents (Begich, Pryor, and Landrieu) also share the distinction of having voted for the Affordable Care Act as Senators, and the law’s mixed popularity could weigh them down, especially among independent voters. Of these three, Republicans stand a fair chance to pick up at least one, if not two, of these seats. Pryor and Landrieu both have long, successful family histories in their states, so they may be able to defend the seats, but it will be difficult.

In North Carolina, the incumbent Democrat, Kay Hagan, is a first term Senator, and also a supporter of the Affordable Care Act. North Carolina went for Obama in 2008 but for Romney in 2012, both in narrow victories. She is vulnerable.

There do not appear to be any Republican incumbents who find themselves in such a difficult position. Susan Collins, Republican incumbent from Maine, is in a pretty liberal state. On the surface, one might assume she would struggle, but, she is fairly liberal herself and no one seriously thinks she will lose.

Of the 29 remaining seats that are being contested, the five most likely to flip appear to be Democrat seats. While other seats will be contested, and surprises could happen (Colorado in particular), the surprises are most likely to lean Republican for now. Why? See below.

National Political Trends

The biggest unknown is how national events will filter into these contests. Issues could arise that may define the race, though for now, it appears that immigration, health care, and the economy will churn through these contests. Immigration is probably the biggest wild card for now, but notice that many of the most interesting races are toward the southern part of the nation. North Carolina, Georgia, Louisiana, and Arkansas may react to immigration dramatically in the next several months. Are those reactions most likely to help or hurt Republicans or Democrats? On balance, given the instability of the southern border, and the Democrats’ willingness to identify themselves as more in favor of relaxed immigration laws, this could play into Republicans’ hands.

More obviously, President Obama’s popularity continues to struggle or decline. Those who disapprove of his job performance now firmly outnumber those who approve. Additionally, and this is the kicker, it is very common for the President’s party to struggle during the mid-term elections of his second term.

Democrats picked up 6 US Senate seats in the 2006 mid-term (when Republican George W. Bush was President) and 8 in 1986 (when Republican Ronald Reagan was President). The only recent exception was in 1998 when Republicans failed to pick up any additional seats during President Clinton’s second term in office. That year was unusual, however, for Clinton was embroiled in a scandal that rallied his own supporters as much as it repelled other voters.

Given Obama’s relative unpopularity, and the historical pattern, it seems quite likely for the Republicans to expect to pick up at least a handful of Senate seats.

Conclusion

What to make of this? The Grand Old Party will, unless things change dramatically, likely pick up at least 4 Senate seats, with a ceiling of probably 7 or 8. That means to control the Senate, the Republicans will need to get close to that ceiling. Given the patterns at work, I think a split Senate, where the Democrats hang on with the VP’s tie-breaker, or a one seat Republican majority seem most likely.

Practically speaking, this will have little effect, except for on the President’s ability to appoint possible Supreme Court justices, which is a big deal. Regardless of the outcome, the President will not sign Republican bills, even if they make it through the Senate, and he will be unable to get his agenda through either house. In most ways, not much will change.