No shortages of predictions for the mid-terms, but a critical shortage of data and foresight
No shortages of predictions for the mid-terms, but a critical shortage of data and foresight

We are now about 90 days from the mid-term election and we can at least start to make some analyses of the prospects.  Will there be a blue wave strong enough to put Democrats in charge of the House of Representatives? And will there be enough Democrat enthusiasm – meaning turnout – to take control of the Senate?  Or, will Trump and the Republicans defy history, the media pundits and the pollsters and keep control of Congress – pulling off another stunning election day victory that throws the #NeverTrump resistance movement into the depths of despair or into the streets in violent protest?

So far, I have avoided any predictions and I am not going to make any now.  Oh sure, it did suggest that the Democrats should not be too giddy in predicting that blue tsunami way back in January.  That was not because I believed they were wrong, but at that early date, they were only dealing in wishful thinking. 

Some brag that they predicted past elections long in advance, but that is not analysis – because there is nothing meaningful to analyze so early on -- too much time and too many events to take place in the interim.  Those prognosticators were simply lucky – and there will always be that group in any election just like some people will accurately predict the winner of a baseball game.

Now that we are at least approaching the event horizon of the black hole of political prognostication, what do we see?

First, there are the so-called historical examples.  According to the popular belief, the party of first-term incumbent presidents generally lose seats in the first mid-term election. The problem with that theory is that it is not entirely true.  Since 1966, the party in the White House has lost congressional seats in EVERY mid-term election with only two exceptions – when Clinton in 1998 and George W. Bush in 2002 gained seats. 

A president’s party losing control of either chamber of the Congress in their first mid-term has only been the case since President Clinton – and even then, it only applies to Democrat presidents.  Clinton lost the House in the first mid-term in 1994 and Obama did likewise in his in 2010. 

Looking at Senate races, Democrat incumbents are running in 25 to the Republicans 8 – with 10 Democrats running in states Trump carried.  Democrats only need to increase their number by two seats to reverse the current 51-to-49 Republican majority.  But, they would have to hold on to every one of the 25 seats to do that – and that is, at this time, unlikely.  It is more likely that Republicans will increase their margin – but that is not by any means a certainty either.

Looking at this another way, Democrats would have to win more than 70 percent of the Senate races just to maintain the current balance.  On the other hand, if Republicans could win just 40 percent of the Senate races, they would have a veto-proof super majority.  It is very unlikely that either party will achieve those results.

What can be said based on the historical record is that when there is a change in party leadership, it is more likely to occur in the House than the Senate.  That is no accident.  That is by design.  Our clever Founders saw the House as the chamber that would quickly respond to the contemporary whims of the electorate.  They fashioned the Senate to prevent a malignant whim from taking over the entire Congress.  That is why only one-third of the senators are up for election every two years.

The more likely Democrat takeover is the House.  Every one of the 435 seats is up for election.  Democrats must oust Republicans in just 23 races without losing any to become the majority party.  That is VERY doable.  When the Obama Democrats took what the president described as a “shellacking” in 2010, Republicans ousted more than 60 House Democrats.

But, what is in that cloudy crystal ball besides historical precedent that can give us some even better clues to the Election Day 2018 outcome?

One of the most reported meters is the so-called “generic ballot.”  Likely voters are asked simply which party you would prefer to control the House.  Currently, Democrats hold about a 13 to 17 percent lead.  Based on past examples, that would indicate a very strong potential for Democrats to take control of the House.  However, the generic ballot is highly volatile.  It was not long ago that Republicans were down by only three or four percentage points.  Major events – and even not so major events – can have the generic ballot going up and down like a seismograph during an earthquake.

There is a bit of a footnote that is necessary when making predictions based on poll numbers.  The GOP generally does better on Election Day – win or lose -- than pollsters predict.  Republicans generally win in close calls – as again seen in the special congressional election in Ohio where Balderson appears to have eked out a narrow victory despite polls showing him losing.  One rule-of-thumb is to take the polling numbers and switch about two or three percent from the Democrats to the Republicans to bring the pollster’s predictions in line with the results.

One of the reasons that the generic ballot may not be a good indicator of the actual outcome of congressional elections across the country is because of the high concentration of Democrat votes in a small number of districts.  It may be true that the majority of American voters prefer Democrats at any given time, but those votes could be concentrated in a few, mostly urban, districts. Hence, Democrats will carry fewer districts by wider margins.  It is the same dynamic that had Trump winning in the Electoral College but losing the popular vote.  Clinton won BIG in California and New York, but that is still only two states.  To use the bridge term – the card game, not the structure -- she had strength, but no length.

Some political tea leaf readers have drawn a correlation between the popularity of a President and the outcome of mid-term elections.  Ever since Democrat Speaker of the House Tip O’Neil proclaimed that all elections are essentially local, that has been the accepted dogma.  Statistics, however, suggest that there is a direct correlation between presidential popularity and the outcome of Congressional elections.

George W. Bush’s popularity was north of 60 percent when Republicans gained House seats in 2002.   The same was true of Clinton in 1998.  Conversely, Obama was hovering around 45 percent popularity when his party took that “shellacking” in 2010.  The biggest losses in House seats occur when presidents are below 50 percent favorable rating.  The worst loss since 1966 was the 2010 “shellacking.”  Obama lost the Senate in 2014 when his popularity had fallen into the 30-percentile bracket.

If this correlation has any validity – and it does appear to have some – the Republicans could be in very serious trouble in terms of retaining House.  Trump’s approval rating fluctuates around the mid-40s.  Based on the statistical correlation, that could mean the loss of up to 60 House seats.

But, the trend lines are never perfect.  In 2006, when Bush’s popularity was cratering in the mid 30 percentiles, the GOP only lost 30 seats – half of what Democrats lost when Obama was at 45 percent.  Of course, if either of the scenarios was repeated, the Democrats would have control of the House.

Some tend to look at retirements as an indicator.  They note that there was an unusually high number of Democrats retiring in advance of the Republican wave election of 1992 – around 40 – and that a similar number of Republicans who have announced their retirement from the House in advance of the 2018 election.  As the theory goes, more than 90 percent of incumbents are re-elected.  The more open seats, the more vulnerability.

There are a number of problems with this correlation.  First, in any given year Republicans tend to retire in greater number than Democrats – and there are years in which this correlation does not exist at all.  The unusually high number of Republican retirements for 2018 may be explained by the fact that a large number were credibly challenged in primary elections.  Others are running for higher office.  Many of the retirees are committee chairs who are term-limited and will lose their powerful posts.  While open seats may seem more vulnerable – and to a small degree may be more vulnerable – the party that had previously held the open seat usually retains it.

Finally, there is the interpretation of all those special elections that have taken place in the past two years.  While political partisans try to bunch them into some sort of trend analysis, they are different events.  Special elections have never been good as presages of General Election results. Democrats did well in the special elections leading up to losing the House in 2010.  Republicans did quite well in the earliest special elections in this cycle.  The first sign of weakness came in Georgia where Republican candidate, Karen Handel, underperformed but still won the election.  The Alabama Senate race that brought down Judge Joe Moore with charges of past sexual misconduct with underage girls was an outlier and should be tossed out of any consideration of trends.  

Democrats broke through with a win in Pennsylvania when Conor Lamb defeated the GOP candidate in a congressional district that Trump had carried by almost 20 points.  Losing the special election in Pennsylvania gave credence of the theory that Democrats were on the rise in Trump country – something that would have to happen if Democrats had any hope of taking the House.  

Ironically, even though Republican Balderson appears to have won the special election in Ohio, the closeness of the race may be the most significant sign that Democrat enthusiasm is a national phenomenon – the making of a wave.  According to some analyses, however, the high turnout in the Democrat counties should have produced a clear victory for Danny O’Connor.  What brought the race back to the Republican side was the high turnout in the Republican counties.

The Ohio turnout – as well as -in all special elections -- can be credited to the unusually large amount of money each party spent, to the on-the-scene big name endorsements and to the amount of coverage provided by the news media.  None of that can happen to that extent in the General Election when there are 435 races – or even in the 50 or so targeted races.

Yes, we can start to look at some trends, but even at this stage, we have to know that the events that will decide this election have not yet happened.  There will be developments in the Mueller election, the Manafort and Cohen trials, new economic news, tariffs getting imposed or trade deals being made, progress or setbacks with North Korea and Iran.  Politics is a lot like basketball.  Throughout the game the lead passes back and forth and each time the respective fans cheer victory.  But, it is only in the final minutes that the game is decided.

Thanks to a cheerleading news media, Democrats have been winning elections every day since 2010 … except on Election Day.  Will that happen again or is this the year that Democrats will break their decade-long losing streak?  Stay tuned.


Audience Index: 23