US Congressional election update – Market uncertainty is the winner

• Republican gains in the Senate slip
• House seats in contention rise
• Forecasts remain for a Republican Senate and Democratic House

The Congressional election has had a volatile trajectory this year. Six weeks ago the Democrats looked to ride the historical advantage for the party out of the White House in non-Presidential elections to a sweeping victory in the House with a takeover in the Senate a distinct possibility. Republican voters seemed apathetic and President Trump’s approval rating stuck in the low 40s promised huge losses for the GOP.

But then came the contentious Supreme Court hearing for Brett Kavanaugh and the media circus that both riveted and dismayed large groups of voters and ended with his confirmation on October 6^th.  

Republican interest soared coming to near equality with the Democrats who had been campaigning against the President for months. President Trump’s approval rating began a steady climb to the mid-40s.  Several of the Democratic Senate seats in states won by Trump moved to a decided Republican advantage. The House decision crept close to balance with the number of projected seat changes just over the needed 23 to switch control to the Democrats.

Advantage appeared to be trending back to historical voting patterns particularly in Senate elections in Mid-West states that were traditionally Republican. But over the past two weeks the election has shifted back to indecision, at least as portrayed in polling.  The incipient trend toward Republicans that was evident in several contests has halted and in a few places reversed.

We look at a summary of the changes in the past weeks then take a detailed survey of the three potential election scenarios in descending order of possibility and their potential impact on markets.

Political predictions

Two weeks ago the Republicans were forecast by RealClearPolitics (RCP) to pick up three seats in the Senate, last week that had dropped to two. With four days until the election it is now one, for a 52 to 48 Republican advantage.

Three weeks ago 29 House districts were in contention at RCP and there were 205 likely Democratic seats and 201 Republican, two weeks ago it was 32 and 205 to 199, last week 31 districts were toss ups and the distribution was 205 and 199. Now the number of House districts considered too close to call is 36 with the Democrats ahead or leaning in 203 and the Republicans in 196 and a predicted 26 seat gain for the Democrats.

Democrats take the House, Republicans retain the Senate

Senate

At FiveThirtyEight, the statistical analysis site, Republican odds in the Senate were 83.2% last week, they have shifted slightly higher to 84.5%. Democratic chances in the upper house were 16.8%, they have dropped to 15.5%.  The average Republican gain is 0.7 seat.

Like last week these percentage shifts have taken place as new RealClearPolitics averages give the Republicans one fewer Senate seat. Last week the Republicans were projected to win two new seats gaining North Dakota and Missouri, while losing no contests resulting in a margin of 53 to 47.  The two Plains states remain Republican this week but now the Republicans lose their current seat in Arizona giving them one net pickup and a Senate at 52 Republican and 48 Democrats.

Of the six Senate seats considered toss ups by RCP, two are led by less than one point, Arizona and Indiana, three, Missouri, Nevada and Florida, have a candidate ahead by two points or less. All five states are within the margin of error for their polling averages and are statistically tied. Only in Montana is there a lead beyond the error margin.

The movement that we saw last week toward Republican candidates in six states, Indiana Missouri, Arizona, Michigan, Nevada, and North Dakota, has been reversed in four, Arizona, Nevada, Michigan and Indiana with the completion of new polls.

Both RCP and 538 award the Republicans one additional seat at 52 to 48.

The current Senate is 51 Republicans and 49 Democrats. 

House

Last Friday FiveThirtyEight Politics, had the odds of a Democratic victory in the House at 84.5% that is now 85.2%. The average Democratic seat gain is 38. Republican odds have fallen from 16.8% to14.8%.

The House averages at RCP last week gave the Democrats a gain of 25.5 seats, two more than the 23 needed for a majority.  This week that addition is 26.  The spread of potential changes was 10-41 seats on October 24th, now it is 8-44, 31 seats were considered toss up then, it is now 36. The change is indicative of greater instability in the election, more seats are in contention and though the average gain increased by just 0.5 the potential for movement is higher if voting patterns favor one party or another.

Of the 36 seats that RCP lists as toss ups, 18 have the Democratic candidate leading, 15 tilting Republican and three tied or with no polling. In the Democratic leads 11 are by two points or less. For the Republican 5 of 15 are at two points or less. If the Democrats won all the seats in which they are ahead that would add to the 203 in which they have solid leads for 223, five more than the 218 majority. Republicans are solidly in front in 196 districts.

The 23 seats that the Democrats need to win control is based on the current House division of 193 Democrats and 236 Republicans.

Markets and elections

Implications for the equity, currency and credit markets are ordered by three factors: the success of the US economy in the past two years; the economic, trade and regulatory policies of the Trump administration; the Federal Reserve’s rate policy.

The outcome of the election will have both direct and indirect impact on the economy and the agenda of the administration. The impact on Fed rate policy is a derivative effect of first two.

One additional consideration is what will be the effect on market psychology of impeachment and the multiple investigations of the Trump administration and its officials that the Democrats may initiate if they gain control of the House?  The key is length. The longer and more intense the questions the greater the potential for a damaging impact.

Market impact - Democrats take House, Republicans retain Senate

This remains the favored outcome, has been so for most of the campaign and as such is largely priced into the markets. A Democratic House and Republican Senate, with narrow majorities in each would lead to legislative and policy gridlock. In the current atmosphere it seems unlikely that the sides would be able to cooperate on anything more than the most necessary funding and operational appropriations.

The administration would not be able to advance its economic approach by making the tax cuts permanent or institute other initiatives but their beneficial effects would remain to encourage growth. Trade and regulatory policies would continue as they are with the potential for an agreement with China available to boost growth.  There is no policy reason that a divided government would damage economic performance.

Currencies

The dollar should remain favored as its economic, rate and policy supports will not be affected by the change of power in the House. American economic growth and the Fed’s rate policies have been the chief factors in that success. Of greater potential impact is the aging of this scenario and new developments in Europe.

The euro and sterling will remain dominated by the outcome of the Brexit negotiation. If, as expected, there is an eventual agreement both will benefit with the larger boon to the pound. The Italian budget will continue to drag on the euro until a face saving solution is found that permits Rome to spend on its economy. Commodity currencies and emerging markets are tied to US and global economic expansion. If the slowdown in China becomes more pronounced it will have a larger impact than the US election. Conversely a trade deal between Washington and Beijing will probably do more to heal growth concerns around the world than any decision of the US electorate.

Equities

The prime concern of equity markets will be the impact of the election on US growth.  Legislative gridlock is historically good for stocks and the economy can be expected to maintain its current pace of 3% and more expansion. With no new policy initiatives the markets will soon be able to answer the question whether the tax cuts and regulatory reform have moved the economy to a long term period of faster growth or simply provided a surge of consumer and investment spending that will subside and take growth down with it. Equites will move accordingly.

Fed policy and credit markets

The Fed’ rate projection of 3.25% in the Fed Funds at the end of 2019 included a decline in GDP to 2.5% next year. The election will not alter the Fed’s view unless the US economic performance changes substantially. The gradual inclination of US interest rates higher rates should continue.

Republicans keep the House and Senate

FiveThirtyEight has this outcome at the smallest odds though with a slight gain from last week.  The Republicans have a 14.8% chance of keeping control in the House down from 15.4% last week. Their Senate chances have improved to 84.2% from 83.2%. Thus their overall average is slightly better though still below the Democratic odds for double control

RealClearPolitics gives the House to the Democrats with an average gain of 26 seats out of 36 in contention. This is but three more than the 23 needed for the Democrats to seize control. A switch of a very few number of seats could keep the House Republican. That possibility is a good deal more likely than the opposite one where the Democrats obtain the Senate thus the higher standing for this outcome.

In an election as close as this the imponderables of motivation and emotion are large and almost impossible to chart.  Democrats have been campaigning for this election for two years, has their passion waned?  Has the Kavanaugh bump for Republicans faded? What effect will the booming economy have? 

The most important imponderable is the same as it was in the 2016 presidential election. Trump’s voters, Mrs. Clinton’s “deplorables’ the ones who put the President over the top in the old industrial Mid-West are not standard Republican partisans. They don’t fit normal voter distribution categories and they are particularly difficult to poll.  Can their vote be transferred from the President to enough local candidates to elect a GOP Congress? That is clearly the logic behind Mr. Trump’s extensive rally tour of contested states and districts. It was the technique that won the Presidential election, will it work again?

Market impact - Republicans Keep the House and Senate

Unified control of Congress would revive Republican plans for the economy. Further tax cuts, another approach to the health insurance market and Obamacare, immigration reform, regulatory reductions and other items will be on the agenda.  Republican economic policies would be viewed by the markets as an assist to economic growth though there will be doubt until proven, as to whether the US economy can expand much faster than its current rate.

Currencies

The dollar would gain modest and probably temporary support if the US economy improves, things in the rest of the world being equal. This will be tempered if the global economy also revives, particularly China. A victorious Trump administration would have a stronger negotiating position in its trade dispute with the mainland and that will make a deal more likely not less. Commodity and emerging market currencies will eventually benefit from the change in scenario from a declining Chinese economy to renewed growth.

Equities

The Republican focus on economic growth would provide stocks with new life. Over time the assumption of better US performance would need to be backed by statistics but stocks would have an immediate and substantial boost from a Republican triumph.

Fed policy and credit markets

The Fed program of normalizing US interest rates would be solidified by Republican retention of the House and the possibility of enhanced economic growth. The governors are not raising interest rates to calm an overheating economy or to forestall near-term inflation but to return rates to a more historical relationship to the economy. Chairman Powell and the board intend to provide a cushion for rate reduction in the next recession. As such the greatest risk is slower economic growth that makes rate increases too risky.

Democrats take the House and Senate

Democratic odds in the House from FiveThirtyEight have improved to 85.2% from 84.5% but their chances in the Senate have fallen to 15.8% from 16.8% last week.  Their odds of a double takeover have fallen slightly since last week.  

RealClearPolitics agrees on the House with 203 Democratic likely seats, 196 Republican and 36 toss ups. The average gain of 26 seats would give the Democrats them 229, a majority of 11.  On the Senate side the Republicans have lost one seat to 52 but remain the favorite.

Market impact - Democrats take the House and Senate

The immediate market reaction would be negative as Republican economic policies which have improved growth, employment and wages would be threatened. It would be markedly harder for the administration to fend off Democratic initiatives without the aid of at least one House of Congress and the President’s international position would be concomitantly reduced.  Beyond the initial impact much would depend on the path chosen by the Democrats.  Confrontation, investigation and perhaps impeachment could be very damaging to economic growth. If the US suffered from the political atmosphere in Washington that could easily impact growth around the world.

Currencies

The dollar would be penalized almost immediately on the assumption that the conflict between President Trump and a Democratic Congress would paralyze policy in Washington and could quickly begin to impact the economy.  This effect would likely rule all of the trading relationships of the US currency and at least temporarily take precedence over local considerations.

Equities

The worst outcome for stocks, immediately lower on reduced economic expectations. Longer term results would, as in the general market view, depend on the governing path chosen by the Democrats. Confrontation would be punished and cooperation rewarded.

Fed policy and credit markets

 A double victory for the Democrats would also be the most problematic for Fed policy.  The tightening cycle requires an economy growing fast enough to support the increases without stalling.  The impact of the larger Democratic sway over economic policies would take time and it might not be negative, depending on the choices, but the immediate assumption would be reduced economic growth. The Fed might be reluctant to highlight its concern by declining to hike in December but until the effect of the political confrontation in DC on the economy was known, policy would become more cautious. The chance of three 25 basis point increases in 2019 would drop sharply.

Statistical methods, RCP and 538

The 12 seat difference between RCP’s 26 seat average gain and 538’s addition of 38 in House races probably stems from the difference in their methods. RealClearPolitics averages separate surveys conducted by independent pollsters. Assumptions about the composition of the electorate, voting habits and other inherently subjective factors for each survey tend to cancel out through the averages. 

FiveThirtyEight Politics builds a statistical model for the country that it employs in all contests. Whatever assumptions about the electorate and voting rates are behind the model are uniform across the nation. To this observer the RCP method better suits the large diversity of political economic and cultural factors in a US election.  FiveThirtyEight’s statistical technique produced a final 90%+ chance of a Clinton victory the afternoon of the 2016 election.

Polling difficulties

One aspect of the election has become increasingly obvious. State and local polling in close races is difficult and unreliable. We only have to look to the 2016 presidential votes in Michigan. Pennsylvania and Wisconsin which were expected to be won by Clinton in all major polls and went for Trump.  

Part of the difficulty lies in technology. How to reach a representative sample of a particular electorate? The ubiquity of cell phones surprisingly makes surveys more difficult as caller id and contact lists make unknown callers far less likely to be answered.  For large groups of voters a phone call is no longer an effective method of polling.

Some pollsters have tried web based or app based surveys but these have different problems from selection bias for the sites that offer the surveys to the political affiliation of young people who routinely use phone apps and the motivation of folks who elect to take web election polls. Others have used a mix of different access methods to approximate their chosen electorate. But whatever the means it has become much more complicated to reach a chosen sample in the last 15 years.

Another difficultly is the application of national surveys to local elections. A large part of the expected swing away from Republicans in House elections is based on the supposed distaste for President Trump among suburban voters. It is true that in national polls his disapproval in the suburbs is high. But a nationwide survey does not necessarily tell you the skew in a particular suburban district in the middle of the country. Many national polls on the popular vote totals in the 2016 election were close to the final results but were wholly inadequate in state elections where the electorate did not mirror the nationwide one. That problems is compounded by the number and small size of the House districts in contention.

Senate polling in some states has been noticeably inconsistent. In Arizona two polls conducted four days apart show a 13 point change in the electorate from +7 Republican to +6 Democrat within a single week. In Indiana the swing was 11 points and in Nevada 10 points between subsequent polls.

Not all states have seen such polling variations. Tennessee and North Dakota have maintained strong Republican leads with consistent and similar poll results for the past three weeks.  Florida has kept a small but uniform tilt toward the Democrat.

At this stage of an election, with forecast changes dependent on a small number of new polls, whose internal dynamics are unknown and with no clear overall trend in sentiment it is the uncertainty that has grown, despite the confident percentages of pollsters.

Generic vote

The spread of Democratic preference over Republican this week is 7.5%.  Last week it was 7.6%.  This difference has fallen two percentage points in eight weeks. Since the beginning of September the  gap has varied from 9.5% on September 4th to 6.6% on October 7^th and 8th, the  days immediately after Brett Kavanaugh's confirmation to the Supreme Court.  The current difference does not indicate a large surge of Democratic voters.

The poll charts the stated desire for a generic Republican or Democratic candidate. There are more registered Democrats than Republicans and this poll has historically overstated the Democrat's predicted vote over the actual percentage by about 7.2% in post war elections. In practice that means a Democratic lead in that area likely indicates an electorate nearly tied.

FXStreet US Congressional election coverage

Republicans gain in the Senate, the House stays Republican - October 28, 2018

Why the House vote is difficult to predict -  October 22, 2018

Is geography destiny? - October 19, 2018

Can the Republican economy hold off the Democrats? - October 14, 2018

Political map - Senate

Political map - House

Political  map - generic vote