Home » Uncategorized » Massachusetts Senate Special Election Round Two: Scott Brown Strikes Back

Massachusetts Senate Special Election Round Two: Scott Brown Strikes Back


Blog Stats

  • 29,050 hits

Start here


Anyone following the political news these days has no doubt heard the reports that Senator John Kerry of Massachusetts will be appointed the next Secretary of State once Hillary Clinton steps down.  With Kerry at the helm of the Foreign Relations Committee, and a well-respected voice on manners of diplomacy, Kerry is a strong choice and easy shoe-in for confirmation (don’t be surprised by a 100-0 vote).  For Kerry, a stint as Secretary of State for the remainder of Obama’s Presidency would be a nice feather in the cap of his career, and complete redemption for his tough loss to George W. Bush in 2004.  Kerry clearly wants the position, is qualified, is respected by Republicans and popular with Democrats, and overall this is a great thing…


Massachusetts Values (and laws)

Normally, such a move for Kerry would be cause for quick celebration for Massachusetts Democrats.  While this means losing their senior senator, their Governor, Deval Patrick, is a Democrat who would for sure pick a Democratic replacement.  In most states, Governor’s have full discretion over their choices (regardless of party) for replacements in these vacancies.  However, the law in Massachusetts is fundamentally different.  Back in July of 2004, the Democratic Super-Majority legislature of the state passed a law that would bar the governor from making appointments in case of a vacancy.  Democrats feared that if Kerry, the Democratic nominee for President at the time, won the 2004 Presidential Election, then Governor Mitt Romney would appoint a Republican replacement.  While a simple solution would have been to change the law like other states have done to require the appointment come from somebody in the same party as the Senator leaving the post, the legislature instead aimed to require a special election for all Senate vacancies.  The bill was vetoed by Romney but overridden by the legislature.  Several years later, when Senator Ted Kennedy’s health was fading after his battles with cancer, a move was made to change the law (after his passing) to allow for an interim appointment between the vacancy and the special election that would take place in between 145 and 160 days from that time.  The new law now allows for an interim appointment, but still a special election is needed.  This of course led to the infamous 2010 Special Election where Republican Scott Brown won the seat of Ted Kennedy.   The race was notable in being the first time a Republican was elected to the US Senate from Massachusetts since the 1970s.

Scott Brown would go on to lose in a general election match in 2012 against Elizabeth Warren at a time when Barack Obama was on the top of the ballot.

However, with the likelihood of a Kerry appointment, Scott Brown already has a chance at a comeback.  Brown is conceivably the only Republican with a strong chance of winning the US Senate Seat, and has ample reason to try for the seat and continue a Senate career.  Brown posts high approvals in the state (his biggest problem in 2012 was that Warren also had good approvals and control of the US Senate was a factor in how the Democratically-aligned electorate cast their ballots) and does not need to worry about Barack Obama being at the top of the ballot.  However, Brown must also know that a win in the special means another race in the 2014 General for a full term and a Republican label will always been a drag on his chances, especially for federal office.  While Brown could mull to lay low and aim for the open Gubernatorial Election in 2014, its a strong likelihood he makes another play for Senate.

Democrats have many candidates, some higher-level than others.  Most of the congressional delegation is no doubt considering a run, as well as Ted Kennedy’s widow, Vickie; other members of the Kennedy family, and Governor Deval Patrick.  Without a clear Democratic front-runner  the focus for now will lie on the state of politics in Massachusetts and the strengths and weaknesses of Scott Brown.  In doing so, we can see where Brown is strongest in the state and weakest; as well as what drives the votes toward Brown unlike other Republicans before.

Politics of Massachusetts

Massachusetts is known to be one of the most Democratic and liberal states in the country.  While it still has elected Republicans to state offices (much like how Wyoming had a Democratic Governor from 2002 to 2008), at the federal level Democrats dominate.  No Republican had been elected to the US Senate from the state sense the 1970s, and the congressional delegation has been completely Democratic since 1996.

Looking at the political fluctuation of Massachusetts can be difficult on the county level.  There are only 14 counties in the state, which can make it hard to analyze in a meaningful way.  However, the state is also divided into 351 towns or cities.  There is no unincorporated land in Massachusetts; everything falls into a town or city.  These towns and cities will be how we judge the partisan nature of the state.

First, lets look at a very crucial detail about Massachusetts.  While the state is indeed very Democratically-aligned, and very liberal, its party registration tells an entirely different story.  In fact, Massachusetts boasts the largest percentage of independent voters than any other state.

Party Reg

Independents are over half the state, and dominate most of the towns and cities, while Democrats, the second largest group, hold in Boston/Cambridge and other scattered cities.  Republicans, all the way down at 11%, don’t dominate in any region.  While the independents dominate, they clearly have a Democratic lean in most races, and even without, Democrats are still well positioned overall.  If Democrats were able to keep 100% of their members in line with a candidate, they would only need 25% of independent voters to back them to still win.  For a Republican to win, they need to not only win independents by a large margin, they also need to win a good amount of Democrats.

  • For example, when Brown lost to Warren in 2012, he still won Independents with 59%, and won 11% of Democrats.  With all that, he still lost by 7%.  In his 2010 win, Brown won 64% of independents and 20% of Democrats to get his 2010 win.

Of course, looking at this map its important to remember that the population is hardly spready out evenly across the state.  As would be expected, a large chunk of the population is in the Boston area.  The symbol map below shows the population of each city, with larger circles representing bigger populations.

Reg Density

Population is largely clustered in the Boston region and its surrounding areas on the coast, with suburban towns and smaller urban centers standing out across the map.

Partisan Leanings in Massachusetts

To asses the partisan politics of Massachusetts, its important to look specifically at the more competitive and high-profile races than the typical blow-outs for Democrats.  Senators Kennedy and Kerry always won overwhelmingly, and many cabinet level Democrats rarely faced strong opposition.  In these non-competitive races, Democrats can come close to winning every single town or city in races with weak Republican opposition.  An example is Ted Kennedy’s last Senate win from 2006.

2006 Senate

Kennedy, in taking 70% of the vote, won all but 3 of the 351 cities and towns.   However, while this shows the apex of Democratic strength, it is highly unlikely to be the case in the special election when Scott Brown is running.

Examined Races

For an analysis of competitive races, there were a few key choices for selection over the past 10 years.  The races chosen represent close-calls, semi-competitive elections, and Republican wins.

  1. 2010 Senate Special Election — This is of course the top of the list, as it was single-handily the  most shocking election results in the state’s history.  Scott Brown beat Democratic Attorney General Martha Coakley in an election that should have been a walk for the Democrats.
  2. 2012 Senate General Election — This race showed the limits to Scott Brown’s appeal.  Brown was challenge by Elizabeth Warren, a Harvard Professor and adviser to President Obama, who proved to be a powerful campaigner and fundraiser.  With President Obama on the top of the ballot and a majority of the state desiring a Democratic Senate, Warren won by 7%.
  3. 2010 Gubernatorial Election — Nine months after the special election shocker, first-term Democratic Governor Deval Patrick was looking for re-election.  Patrick had a rough first year, and while his approvals has bottomed, they were rebounding slowly as the election drew.  Patrick was at risk to a Republican moderate challenger (Baker) and a challenge from conservative Democratic-turned Independent Tim Cahill (the state Treasurer).  Patrick won re-election by 6%.
  4. 2002 Gubernatorial Election — The second Republican win, when Mitt Romney won the governor’s race in a close race.  Romney benefited from a Democrat weakened by the primary campaign.
  5. 2006 Gubernatorial Election — In 2006 Romney was not popular and risked losing re-election.  With plans to run for President in 2008, Romney opted not to run and left the party trying to hold the seat in a year that was looking to be very favorable to the Democratic Party.  Deval Patrick, after winning a crowded primary, went on to win the race by 20% (with 55%, and 10% going to assorted independents), in a race that could have been competitive but slipped away from Republicans in the blue year.

The following races can be seen in the maps below.  Pay attention the the color shadings (which are uniform for each map).  Notice how much more dark red exists in Scott Brown’s win (representing larger wins). 2010 Sen2012 Sen 2010 Gov 2002 Gov 2006 Gov

Looking at the first three maps show a pattern of Republican support in key sections of the eastern half of the state.  The key to Brown’s victory was to really rack up the vote in those areas; while Romney’s win came from a more scattered win across the state (granted that race was 10 years ago and regional strengths have since shifted).

Based off these results its possible to create a party loyalty map based on the results.  I broke down each town into a category based off its results in the competitive races.

  • Strong Democrat — Democratic wins in all five races
  • Likely Democrat — Democratic wins except in 2002 or 2010 Special Senate Election
  • Likely Republican — Republican win except 2006 Governor
  • Strong Republican — Republican win in all of the five races
  • Swing — Variation wins depending on the race

Party Loyalty

Compare this map with the registration density map, and you see that overall Democratic towns hold much more population than Republican areas.  For someone like Brown to win, he had to win the Likely Democratic towns.  Democratic areas hold 2,239,652 voters, while Republican areas only hold 1,207,651 registered voters.  The swing region isn’t enough to close the gap (just over 866,000 voters), so any Republican win involves eating into the Democratic areas.

The situation is difficult for Republicans to navigate.  Looking at Scott Brown’s 2010 win will give a better look at a Republican path to victory.

Scott Brown

In the winter of 2009 Scott Brown was just a state senator in Massachusetts.  With a reputation as a moderate, he was the best choice for Republicans who is largely believed they had no chance of winning the US Senate seat.  Brown won in a major upset, and instantly became a household name.  Brown now finds himself a popular figure in the state despite losing re-election.  He was defeated while holding a 60% approval, to a 32% disapproval.  His loss was thanks to Warren’s popularity and the state’s wish for Democrats to control the Senate.  Brown’s image is now that of a moderate, acceptable statewide Republican (deserved or not is not the issue).  This makes Brown a strong candidate for the election.

Brown’s 2010 win should be looked at even closer.   First, lets look at the actual percentage breakdown by each town.

Brown 10 percent

Brown dominated in middle and eastern Massachusetts (minus the urban areas around Boston), winning over 60% of the vote in many areas.

Compare this with the percentage map for Charlie Baker, the 2010 Republican Gubernatorial nominee would lost by 6 points to Deval Patrick (but this makes him the most successful Republican candidate except for Romney, which was 10 years ago).  Baker and Brown won similar areas, but as the map below shows, Baker’s percentages were smaller than Brown’s.

Baker 2010 percent

Baker won many of the same towns as Brown, but one them more with 50%+ rather than 60%+.   Brown’s winning wasn’t just winning so many areas, but the sheer margins he racked up where he did win.

Lets look at this further.  Lets look at how much better Brown did in 2010 than the Republican nominee for Governor did in 2006 against Deval Patrick.  The map below shows the percentage improvement Brown had compared to the 2006 election for Governor.  Higher percentages mean more of the vote Brown got compared Democrat Deval Patrick.

Brown gain

In several areas Brown gained over 25% of the vote compared to his Republican counterpart from 2006.  Brown’s increases coincide with the big wins he had in the middle and southeast of the state.

Look back to the party loyalty map, and lets see how Brown did with the towns/cities that fell under likley Democrat and swing.  Did Brown win most of the swing towns?  Did he rack up alot of the likely Democratic areas?  The map below shows yes.

Brown 10 crossover

Brown overwhelmingly won the swing towns (dark red), taking almost all of them.  Meanwhile he also won several likely Democratic towns (light red).   He had strong appeal with the swing region and a good amount of crossover appeal in 2010.

However, Brown was not able to use his crossover appeal and good approvals to ensure re-election in 2014.  Warren beat Brown by 7%, thanks to her taking back ground that Brown had gained in 2010 for the Republicans.  The map below shows the losses in vote Brown suffered between 2010 and 2012.  Positive numbers are losses for Brown, and negative numbers are actually Brown performing better (the numbers were achieved by subtracting Brown’s 2012 percentages from his 2010 percentages).

Warren Gain

As seen, Brown lost ground almost everywhere except a few scattered areas.  Warren made decent gains compared to Coakley in the southeast and middle Massachusetts, blunting some of Brown’s margins (even though he still one many of those areas).

Looking at Brown’s 2012 loss on the party loyalty scale, we see Brown still did well in the swing towns, winning almost the same amount, but the key difference being he didn’t win a single likely democratic town.

Brown 12 crossover

Brown still did well in the swing areas (as demonstrated by him winning independents in the exit polls).  However, in Massachusetts its not enough to take the swing region.  Without crossover into the Democratic towns and cities, Brown couldn’t win re-election.

Despite Brown losing ground in 2012, the biggest losses came in shrinking margins, not a conceding of entire towns/cities.  Brown won many of the same areas, in both races; with 2012 simply seeing him get smaller margins and losing likely democratic towns.  However, the map below shows little variation in how towns/cities have voted overall when Brown is on the ballot.  Largely towns have either always voted for Brown or never have.  There is a handful of towns that defected from Brown, and just a few that moved to Brown.

The map below shows which towns always or never voted for Brown in his two senate campaigns.

Scott Brown

The map shows a pretty clear divide, with only a handful of towns switching sides between 2010 and 2012.  Warren’s pickups were significant, as the light blue towns represent over 500,000 registered voters.  However, the margin of the win also came from eating into Brown’s margins in the cities/towns he always won.

Overall there is a clear divide in any election where Brown is on the ballot.  Independent dominance means that an election can swing either heavily to Brown or against, depending on issues and candidate appeal.  Perhaps though, there is another metric to help determine the likelihood of a city/town voting a specific way in a competitive race.  For that, I look at something other than partisan strength and registration.    For this I will look at religion:  specifically Catholicism, in Massachusetts.

Catholicism in Massachusetts

Massachusetts has the reputation for one of the most liberal states in the nation; but it is also the most Catholic.  The state is around 54% Catholic.  While this distinction clearly means that the Catholics of Massachusetts don’t fall in line with Catholic social doctrine (which is much more conservative), the Catholicism of the state can still give an insight into how moderate Republicans can make gains in the state by appealing to the religious sensibilities of the voting populous.

Religious data can be harder to come by that another demographic statistics.  Religious affiliation is not part of the US census, and is rather collected from he American Community Survey and the works of religous groups. There is no town-level data, but there is county level.  The map below shows the counties of Massachusetts and the percent of them that is Catholic.

Catholicism in Mass

The most heavily Catholic counties are in the southeast, the same areas that voted for Brown.   However, the Senate races are not the comparison being looked at for this. Rather this section looks at a referendum that took place in 2012.

Referendum 2 was on the ballot right alone with the Brown v Warren senate race.  The referendum would have legalized doctor-assisted suicide for terminal patients who requested it.  The measure was HEAVILY opposed by the Catholic Church, which invested money and manpower into battling against the measure.  There was a divide in the state, but the measure could have likely passed if the church had stayed out of the race.  Instead, the measure failed 49% to 51%.

Ref 2 by County –

There is a significant correlation between the vote against the referendum and the catholic population of the counties.  The scatterplot below shows that as the Catholicism percent for the counties increased, support for the referendum fell.

Catholic - Referendum 2 Scatter

So we see that Catholicism and referendum vote had a correlation to each other.  While I do not have data on Catholicism by town/city, I do have the results of the referendum by town/city.

2012 Referendum

After seeing several maps of Republican strength (and especially Scott  Brown strength), this map is sure to look familiar.  The referendum failed in the same areas Brown and other Republicans have polled well, the southeast, parts of the northwest, and the middle of the state.

Lets compare the 2012 referendum vote to the 2012 Senate vote.  Warren obviously had to win areas were the referendum failed, and Brown won some areas were the refernedum passed.  However, both candidates largely had marginal crossover when it came to the referendum.

2012 Senate and Ballot 2

The red towns are ones that voted for Brown and for referendum 2, while the blue voted for Warren despite voting against the referendum.  In total, 101 towns/cities crossover in one way or the other out of the 351, so less than 1/3.  Over 2/3 of cities/towns had the same preference for Senate candidate and the referendum, showing a correlation between the referendum vote and the vote for Senate.

The scatter-plot below shows that as support for referendum 2 decreased, support for Scott Brown increased.  The correlation is not overwhelming but still significant, with several outliers, but an overall trend.

Brown Ref 2 scatter

The trend is visible, but there are many outlier towns that voted against the referendum but still for Brown (showing Brown’s crossover) as well as anti-referendum towns that voted for Warren.  But the bigger trend is visible as towns that start in the upper left and work there way down to the center right (signifying an increase for Brown as support for the referendum fell).

The same scatterplot using Scott Brown’s 2010 percentages shows an even better correlation, however.

brown 10 ref 2 scatter

There are fewer outliers this time, and a more clear slope.  Due to Brown winning more areas, towns that voted against the referendum but still voted for Warren didn’t exist.  Rather, Coakley had less appeal in the anti-referendum areas.

Brown isn’t the only one with a correlation to the referendum vote.  Republican Gubernatorial nominee for 2010, Charlie Baker, also saw his support increase as referendum 2 vote decreased.

baker ref 2 scatter

The scatter plots and maps show a correlation between Catholicism and the referendum vote and the referendum vote and partisan appeal (specifically with a strong republican like Brown).  Thus, Catholicism itself can be a possible indicate of partisan leanings in a competetive race.  This of course does not mean any Republican running has a basement of support, or a Republican can be very conservative.  Rather, it indicates a moderate Republican can have appeal with the independent, and Catholic voters who are liberal leaning but considering they vote Democratic in most races, feel very comfortable voting for a moderate Republican like Brown to satisfy their moderate and sometimes right-of-center leanings on specific issues.


Overall, Massachusetts is, and will continue to remain, a Democratic stronghold.  Republicans will perform better in gubernatorial and state cabinet races than federal races due to the overarching willingness of people to cross party lines on officials they are not sending to DC to cast votes.  While Romney and Baker had some statewide appeal, they would likely be crushed in federal races for Senate.  Scott Brown offered a unique appeal that crossed into federal elections.  Brown could likely win the 2014 Gubernatorial Election, but in opting for the Senate seat, his chances are less than with the governor’s race, but still even money (and better of a battered Democrat emerges from the primary).  Brown winning would come from big wins in the Catholic areas (perhaps campaigning on government overreach dealing with contraception or religious freedom).

Brown’s strengths lie in the southeast, middle, and parts of the northeast.  Democrats do well in the west and in the Boston metro areas.  Democrats need to work to hold into the towns/cities that Warren won in 2012 that Coakley did not.  In addition, tamping down Brown’s margins in his strongholds will also be important to diminishing his lead.  Campaign strategy should focus on the need to keep the Senate seat in Democratic hands.  However, Democrats need a strong and likable nominee.  It is not good enough to simply focus on a strategy of “you like him/her, but they are of the wrong party” (Republicans tried that in the 2012 North Dakota Senate seat and lost).  Democrats will need a strong and credible challenger.  The primary campaign, if it indeed becomes a major battle, must stay positive, with Democrats not attacking each other.  A primary campaign could help Democrats build the name recognition of their candidates; provided it stays positive.  While the primary goes on, the DNC and DSCC need to be involved, building a ground infrastructure to boost Democratic turnout, and if needed hitting Brown for controversial votes.  Let the Democratic candidate remain positive as long as possible.

Part of the problem in 2010 was Democrats sleep walking the race.  By all accounts are they will not make the same mistake this time.  Scott Brown is a strong and formidable challenger, but with Obama pledging to campaign for the Democratic nominee Brown may not have the pull to get over 50%.   Democrats have a good shot at holding the seat, but mistakes or complacency cannot happen.  Scott Brown is a formidable candidate and can easily replicate the events of 2010 if Democrats are not careful.

UPDATE:  Congressman Ed Markey enters

Sense my original post, congressman Ed Markey has entered the Democratic primary for the senate seat.  Markey has been a member of the house since 1976, representing the areas north of Boston.  Markey has rarely faced a serious challenge and is a high-profile liberal member of congress.  Markey’s signature issue has been environmental causes, and he was the co-author of the 2009 Cap and Trade Bill that passed the house.  Markey’s role as an environmental advocate parrells with Elizabeth Warren’s role as an advocate for consumer issues and banking regulation.

Markey has already received the endorsements of the Democratic Senatorial Campaign Committee, John Kerry, and Vicki Kennedy; making him the clear front-runner in the Democratic primary.  Markey will be a strong candidate against Brown.  He has routinely won nearly every city/town in his recent re-elections.

2010 7th2012 5th

Its not that Markey’s districts are democratic vote-sinks either.  The congressman’s district (old and new) has held towns that voted for Scott Brown in either 2012 or 2010.   Overall, these towns which voted for Brown also gave their votes to Markey.

Markey Cross-over

The map above shows several towns voted for Markey in either 2012 or or 2010 while also voting for Brown at different points.  Only one town in his district, Lincoln, has shown a willingness to vote against Markey and vote for Brown (granted it voted for Markey half the time).

Markey represents a strong candidate for the Democratic Primary, where his liberal credentials will make him a strong favorite.  His environmental activism gives him a strong biography for the liberal state.  In a general election matchup against Brown, he stands a strong chance of prevailing.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: