What a divided US government could mean for policymaking, the economy, and markets

Paul Diggle: Hello and welcome to Macro Bytes the economics and politics podcast from Aberdeen Investments. My name is Paul Diggle.

Lizzy Galbraith: And I'm Lizzy Galbraith.

Paul: And it continues to be all go in the world of macro and markets. The bond market, our topic of a couple of weeks ago on the podcast is very much in focus. But there is a big macro event on the political calendar looming. The US midterm elections on the 3rd of November, about a month from when we are recording this. And President Trump is on the ballot, whether he likes it or not and despite the fact that this is a congressional and gubernatorial election, because midterms are usually a referendum on the sitting president and the Republicans are facing many of the same headwinds that have hurt the most incumbent administrations in recent decades. So I'm going to be asking Lizzy all about that, what it means for the economy and markets and what the results of the midterms may tell us about the 2028 presidential race. So Lizzy, let's get into it. Opening question, how would you characterize the state of the race heading into the final few weeks of the campaign?

Lizzy: So I think we're seeing growing democratic momentum in the polls at least. We're seeing quite a lot of voter dissatisfaction with President Trump, particularly his handling of the economy. And that looks to be largely stemming from pretty steep increases in energy prices over this year. They're slightly pulling back from sort of all-time highs, but they're still really very high, particularly gasoline. And that is creating a situation where Democrats actually look to be a lot more competitive in states that we would consider to be, under normal circumstances at least, pretty safe Republican seats. So we don't actually have, I think, the midterm race that we had initially thought we were getting when we were looking at this, you know, at the start of the year and sort of preparing what we thought was going to happen. It's actually turning out to be a slightly different race, certainly a more competitive race in the Senate in particular. 

Paul: So we've put out a forecast for the results. Clients can access it on our website. We think the Democrats win the House. Most people, almost everyone, I think, thinks that. And we forecast that the Democrats will win the Senate 51-49, but it's a close-run thing. And on that note, Lizzy, you and our colleague Bob Gilhooey run a model to forecast the outcome and then you overlay it with a series of judgments as well. So why don't you tell us a bit about what that election prediction model is saying, what goes into it, how does it work?

Lizzy: Yep, so the model as it stands gives Republicans a 52% chance of holding on to the Senate. We then overlay that with the polling judgments that we have and we can see from those polls that there's a bit more Democratic strength in some states where the model is maybe focusing on the fundamentals. That's a deliberate choice. The model is a fundamentals-driven model, not a polling-driven model to try and cut out the noise that you sometimes get, particularly in US elections. There's a lot of polls, not all of them always turn out to be particularly accurate. So we've taken a deliberate sort of two step approach to thinking about the US election. Step one is this model, which is a race level fundamentals only Senate-driven forecast. The only polling that we actually use in the model at all is presidential approval rating, because we can take that all the way back to the 1970s and actually further if we wanted to. We then layer on top of that polling judgments to get our final race call for each. So at the moment, although the model is saying 52% likelihood of Republicans retaining the Senate, our actual call is for Democrats to flip the Senate, albeit by a very narrow margin. We do still put quite significant weight, I think, on the odds of Republicans retaining that majority as well. But as of this week, the official forecast is for Democrats to take both the House and the Senate on November 3rd.

Paul: When you say fundamentals-driven model and you distinguish that from the polling models and it's polling models that sort of get a lot of the limelight when you see headlines about political prediction models in the US, they're often polling-driven, fundamentals-driven is a different approach. What are those fundamentals? What are the actual variables in the model? Why are those good election predictions? 

Lizzy: So there's four buckets in the model, I think we could say. So the first is partisan lean or partisan inertia. So the likelihood that voters in each state just vote with the same party every time, and therefore you have a different sort of baseline for change in each Senate race. Some states are going to be more purple than others, more inclined to have a large proportion of swing voters. Some have voters that do just habitually vote with the same party every time. So that helps us gain a bit of a baseline assumption about the state of play in that state. You then have the characteristic of each race itself. So is there an incumbent? Is it a midterm or a presidential? Because that affects things like turnouts and also the types of voters that vote as well. State level economic conditions has become a very important part of this model as well. So employment levels, energy prices of course. So again, to try and capture that variance across the US in terms of the conditions that voters are experiencing. And then finally, national level sentiment as well. How unhappy or happy are voters generally across the country at the moment? And that one is a pretty easy answer. Voters, generally speaking, are not very happy in the US at the moment, consumer sentiment is fairly poor. And as I said at the start, voter satisfaction with President Trump himself is also fairly poor. 

Paul: Are the midterms then best thought of, as I said at the open, as a referendum on President Trump or are there broader factors at work that are affecting this likely, you know, big swing towards the Democrats?

Lizzy: So there's definitely a midterm penalty that the party of the incumbent president tends to pay at a midterm. We would always tend to see the party of the president lose ground in a midterm election. They tend to lose both House and Senate seats. Now, this time, we're dealing with a particularly negative set of circumstances for the party of the president, because on top of that midterm penalty, you also have economic conditions that are weighing on Republican chances. Diesel prices in particular seem to be really negatively affecting Republican voting intention, even it states that we would normally describe as having a fairly strong Republican partisan lean. That appears to be exacerbating that existing habit of voters to move away from the party of the president at the midterms. So this time I think it's a bit of both. Republicans are always likely to lose ground here, but whereas at the start of this cycle, we were looking at a Senate map that actually contained a lot of structural advantages for Republicans that we thought were going to help them retain the Senate, now it actually looks like some of these additional factors might mean that Democrats are actually able to overcome some of those disadvantages because voters have moved quite strongly away from the president.

Paul: Let's get to the Senate map and individual Senate races in a minute. But tell us about the House first, because that seems like a much clearer cut outcome. So what is the scale of Democrat victory in the House that we're expecting? And why is there quite a lot of confidence in the forecast that the House is going to flip to the Democrats?

Lizzy: Yeah, so the House is actually a much easier call to make in a lot of ways. It's easier to see the House as the closest to an accurate representation of the mood of the country at any given time. All seats are up. Generally speaking, it's the one that is most responsive to voter sentiment of the two chambers. So it tends to be quite swingy. You do tend to get a lot of churn in elections. And it's also notable that Republicans are sat on an absolutely razor thin majority already. So it really doesn't take much to see the chamber flip. And given that we're looking at a national picture where generic congressional voting intention is a sort of D+9 type environment at the moment, on average, that is more than enough to see a swing away from Republicans, clear that very, very small hurdle actually to secure the majority. And even with gerrymandering, where states have changed the congressional map to try and input some type of partisan advantage into their maps, we're not seeing any indication that that's done enough to protect the Republican majority or even actually give the Democrats a particularly large majority either. So we're still expecting actually the swing to be fairly small because you've seen the number of competitive seats decline over time because of that gerrymandering effect. Even still, a majority of between 10 and 30 I think is fairly likely. It all depends on how accurate these generic congressional voting intention polls actually are. D+9 environment would indicate a relatively larger swing within that kind of 10 to 30 bound. 

Paul: So the Senate is much closer than that, though. As I said, our forecast with the model and then overlaying judgment of individual races and taking some information from the polls is currently 51-49. So why is that race so much closer? And tell us about some individual Senate races that are really going to be critical to determining control of the chamber.

Lizzy: So first of all, it's closer because this Senate map, the Senate only has a third of its seats up for election every cycle. So there's always a bit of change in terms of the states and exactly which seats are up each election. And this particular cycle looked to be pretty advantageous for Republicans. They had a lot of safe seats up. Democrats, by contrast, were defending more seats. and the seats that they would need to flip to actually take control looked to be pretty out of reach. They tended to be fairly red states. So Democrats need to win four additional seats to take their total up to 51 seats. That clears the vice presidential tie break that gives Republicans a bit of a buffer. At the moment, they don't need to use that, but it does give them wiggle room. If they fall down to 50 seats and we see a Senate that is split down the middle, JD Vance can come in, break the tie in favor of the Republicans, and you still have a sort of functioning Republican majority. Democrats obviously do not have that advantage, so they do need to get to 51 - four additional seats. They also have to defend seats that they already have, some of which are still competitive seats. In particular, you're looking at Michigan on the Democratic side, where that race is still looking fairly close. And the pick-up opportunities for Democrats look fairly challenging in a lot of places. We think that there are some safe bets within those potential races. So North Carolina and Ohio, two states that are currently Republican held, they look like they are really quite likely to flip to the Democrats. You're looking at Democrats in North Carolina leading by over 9 percentage points in polling averages, in Ohio it's closer to five. But regardless, they're well beyond what you would see as a sort of margin of error or a standard polling error. You then have a different category of races underneath that, races that you could all describe as being fairly close to toss up. So, and a lot of those are leaning towards Democrats right now in polling averages. So you have Alaska where Democrats are ahead by almost 3 percentage points. We actually do not have that race called for the Democrats in our current forecast. We think that stays with the Republicans, but polling has been trending towards the Democrats over the course of that race. You have quite significantly, I think, Texas, this race has really captured the imagination of a lot of people. There you are seeing the Democrats ahead by over 3 percentage points. That race we have called for the Democrats because of the durability of that lead now is very narrow, but it has been very consistent over a number of months now. Maine is another Republican held seat at the moment that has been a Democratic target now for a really long time. Maine is an unusual state because it does tend to vote very reliably for Democrats in presidential elections but it has a long-standing Republican incumbent, Susan Collins, who consistently overperforms Republicans in Senate elections and has managed to beat back Democratic challenges multiple times now. So we have that race called for the Democrats, but at the moment it's actually the closest of the calls that we have. Democrats are ahead, but only by two percentage points. So it is very possible that Susan Collins can hold on even in this very negative Republican environment that we have. And then finally, we have this category of sort of wild cards, Senate races, where under normal circumstances you would really I think, be coloring them in red before you'd even begun to sort of hear votes being counted. But they are showing Republican weakness at the moment. So you have Iowa, where Democrats are ahead by nearly two percentage points in polling averages. Kansas where Republicans are ahead by a whisker, only 0.3 percentage points in polling averages now, and Nebraska, which is a slightly unusual race because it's actually not a Democrat or a Republican that's ahead, it's actually an independent candidate against a Republican incumbent, probably would caucus with Democrats if he were elected, but it's a slightly more individualized race there. Now we have all three of those races being called for Republicans at this stage, but it does show the strength of, I guess, the anti-Republican feeling at the moment that we even have those races as question marks at the moment. So all told, our current forecast is for Texas, North Carolina, Ohio and Maine to flip to the Democrats. That gives them 51 seats and counters the Republican vice-presidential tie break.

Paul: So talk to us, Lizzy, about polling errors and how we should be thinking about polling errors. You've explained that we are basing our forecast on a fundamentals-driven model that doesn't use polls and deliberately because of worries about polling error. We are overlaying a bit information from the polls to then sort of aim off that forecast a little bit and how we arrive at our call. Are we expecting large polling errors this time? Is there reason to expect they'll be systematically in the direction of the actual vote favoring Republicans more than the polls would suggest? Why have recent polling errors been in that direction? How should we think about polling errors in this context?

Lizzy: So it's absolutely possible that we see a polling error either in certain states. It is very true that some of the states that we are currently seeing very close polling in, very unusual polling by sort of historic levels, are places where you have seen polling errors in the past, and particularly polling errors that flatter Democrats. So it is completely possible that we are seeing something of an overstatement of democratic strength in some of these states. And of course, we cannot rule out that is a national trend. It's impossible to tell before we get to the final votes as to whether or not that has happened. But it's definitely true that you should you should always have in your mind, I think, that there could be a degree of inaccuracy in the polls. And it is also worth saying that every single one of these polls, for what it's worth, we only ever use averages. We're never ever taking a read from a single poll and then basing a judgment on that. But all polls come with a margin of error attached to them. And that margin of error is usually something between 2 percentage points to 3, depending on the pollster. And with the really narrow margin that we're seeing in some of these races, even just an overstatement within the margin of error of these polls could be enough to create a completely different Senate map to the one that we have ahead of us. So I think there is sometimes a willingness to call all inaccuracy in polls a polling error, whereas sometimes it's actually just that they are within their own margin of error as stated by the pollster itself. So it is important to have a caveat and to think about the polls as potentially being within a range. And a lot of these polls are so narrow that, we don't necessarily have a high degree of confidence in a lot of these states. It is entirely possible that Iowa, as we have called it, is not, is going to be Republican on election night, whereas the polls are currently showing the Democrats with a small advantage. Likewise, it is also completely possible that some of these Democratic states have overstated the support of the incumbent. Michigan is one where, you know, the leader there is still narrow. So I think there is something to be said for having a degree of caution and humility around what these polls are telling us, but not all inaccuracy is necessarily an outright polling error either.

Paul: Yeah, hopefully everyone by this point understands that the health warnings that come with polls and any election prediction modelling and of course our prediction is a probabilistic prediction because there are of course ways in which it could be, say, a split Democrat House Republican Senate. It's very possible. But let's move from talking about prediction of the outcome to implications. And so what? So let's assume the House goes Democrat. As we've said, the Senate's a bit less clear. But either way, what does that mean for policymaking in Washington over the next couple of years?

Lizzy: So if the Republicans lose one or both chambers then the sort of the top line is that President Trump finds it far harder to pursue legislation and increasingly relies on his executive authority. Now, in some ways, that is not that different to the picture that we've seen over the past two years, because, as I said, Republicans have had such a narrow majority in the House actually, they find it really hard to pass legislation and beyond the One Big Beautiful Bill Act, which was passed through reconciliation there hasn't actually been much in the way of legislative action in this Congress. So it isn't necessarily a huge change. And it's worth remembering that President Trump is already fairly reliant on executive orders. He likes using them. He enjoys the immediacy of that style of policy making rather than going to Congress and sort of haggling for what he wants. So in some ways, not too different. But Congress controls the public purse. President Trump cannot get around that. And therefore, if you have Democrats in any way involved in running Congress, you have to negotiate with them for your next two budgets. And that's where the influence really comes in. We expect that a divided government, maybe slightly counterintuitively, actually leads to some modest fiscal loosening because in exchange for the defense spending increases that President Trump has been seeking, his ask is for a $1.5 trillion defense budget for next year - I don't think that was likely even if the Republicans do hold on - but nonetheless, defense spending increase is clearly what this administration is prioritizing for next year. The Democratic priority is for a rollback of cuts to healthcare funding. If you take a slice off of both of those two asks, then you get an increase in healthcare funding in exchange for an increase in defense spending and therefore modest fiscal loosening relative to the current baseline that we have for US spending. Nothing dramatic, but nonetheless not quite the assumption that I think some people hold around US divided government that it is essentially fiscally neutral and that nothing happens.

Paul: Okay, so that's an important nuanced fiscal loosening albeit modest at a time, of course, as we said at the open, of substantial bond market pressure. So what does all that mean for markets?

Lizzy: So I think there's often a preference in markets for a divided government because the assumption is that divided government equals gridlock. And that's far nicer to deal with than, constantly worrying about debates and legislation and who's doing what. So I think on net I'm not sure that people will be particularly worried by the concept of a divided government. And it's worth bearing in mind that even if Democrats do take the Senate as well, that is still a form of divided government because you still have a Republican president with veto power. So just because Republicans lose both chambers does not necessarily mean that this picture looks fundamentally different, actually. But gridlock is still, by and large, aside from your must pass bills like the budget, going to be the outcome here. But true gridlock, absolute silence from the government is possibly not what we're going to get because you're going to have an awful lot of posturing ahead of the 2028 presidential election, lots of politicians in Congress looking to build their profiles. And one thing that Democrats are likely to do is to use the control of committees that they'll get upon entering the majority to launch an awful lot of investigations into the president, into the president's associates, and also probably into sort of really hot topic issues with voters like AI safety, data centers and that could involve bringing in CEOs for hearings. Those could be fairly publicized. So while I think the likelihood of legislative action is fairly remote, this quite public scrutiny of large business leaders, of certain politicians, of people around them could still generate quite a lot of noise and affect sentiment.

Paul: Yeah, that's a really important risk, I think, to watch out for going into the next Congress more, even if we don't see legislative action on reining in AI in some way. The president has made it quite clear that he is not in favor of that. Even just hauling up tech bosses for questioning turns up the pressure and perhaps shows you as what might be in store 2028 and beyond. So I think that's a very important theme to watch. Okay, penultimate question, Lizzy , does it matter if it is split Congress versus unified House and Senate under Democrats? You've said that either way it's still split government with a Republican in the White House, but are there any nuances between those two combinations?

Lizzy: Yeah, so the Senate does have an important role that the House doesn't in terms of its ability to confirm or block presidential nominations. That includes judges, cabinet officials, executive appointments. So it does give it additional leverage in terms of the president's agenda that you don't get if you just control the House. So that ability is solely within the power of the Senate. And it's definitely true that I think the Democrats would have slightly more leverage overall in these types of fiscal negotiations that we're talking about if they control both chambers rather than having to negotiate twice in a way. So while I still think that either combination still means very little in the way of legislative action, you do have a bit of enhanced influence in the full political process if you have the Senate as well. And that ability to make or break nominations, particularly on the judicial side of things, has been strategically very important in previous constructions of Congress. So I definitely wouldn't discount that in terms of the long trajectory of US politics. But in terms of the practical macro implications of this upcoming Congress, I don't think there's too much difference.

Paul: Okay, final question then, what will the results tell us about 2028 and the next presidential race?

Lizzy: So this is one of the reasons that I really, really love midterm elections. And I think they're absolutely fascinating because in my mind, a midterm election is actually the start of the next political cycle. It is not the middle point, because beyond this point, and particularly true of this year actually, because we have two primaries coming up, one for each party. All thoughts then move to posturing for 2028 primaries. So if the results are, as they look on paper, fairly bad for the president, it's possible that his influence over the party actually starts to decline over the next few months. And that maybe means that his endorsement matters less in the Republican primary or you see a wider sort of slate of candidates look to run a more active debate over policy on the Republican side. For the Democrats I'm expecting that primary to be pretty big to be honest because the Democrats are having more free rein I guess over the ideological direction that they want the party to be in and it does appear to be a slightly unanswered question for them at the moment. These Senate races are quite important in terms of setting the tone of that debate because we have some fairly high-profile progressives running in key states. So the success or failure of their runs, I think, really determines whether or not that progressive wing of the Democratic Party starts the 2028 primary cycle on and up or whether it's still really struggling to gain credibility with the rest of the party. So I'd be really looking in that context at the outcome of Michigan and Maine. They're two really important seats for the Democrats, but they definitely have candidates in them that are from outside the centrist wing of the Democratic Party. So that will be a really key debate to watch after we actually get the results in. And then finally, I'd also be looking at any senators, either incumbent or new, that really outperform on election nights. Now the obvious choice here based on the polling would look to be someone like John Ossoff from Georgia. Georgia should have been a swing state and it now looks like it's so safely Democratic we don't even talk about it anymore. But it is a state that President Trump won in 2024 and so he is showing an awful lot of strength in a state that should be really quite purple. So I'd be looking at those senators that can outperform the national environment and see what they do next, whether they look to use that profile to build a potential candidacy or whether they're looking to build influence and build some sort of political coalition to further their particular vision for the party after this point. But it should be really, really interesting after the midterms usually swing immediately into talking about who may or may not run for president next. So I would expect that conversion to be really quite quick once we know the results.

Paul: Great. Well, I look forward to us coming back and talking more about that on the podcast. But that's about all we have time for this episode. Lizzy, thanks so much. Thank you to you for listening. As ever, please like and subscribe to Macro Bytes on your podcast platform of choice. but until next time goodbye and good luck out there.

This podcast is provided for general information only and assumes a certain level of knowledge of financial markets. It is provided for information purposes only and should not be considered as an offer, investment, recommendation or solicitation to deal in any of the investments or products mentioned herein and does not constitute investment research. The views in this podcast are those of the contributors at the time of publication, and do not necessarily reflect those of Aberdeen. The value of investments and the income from them can go down as well as up, and investors get back less than the amount invested. Past performance is not a guide to future returns, return projections or estimates and provide no guarantee of future results.