Do election signs work? #nlpoli

Apparently, they do.

In four randomised field experiments, researchers at Columbia University found that lawn signs increased vote share by about 1.7%. They published results of their research online on Christmas Day.

That isn’t going to win an election single-handedly but it could be the difference in a tight race. The advertising didn’t drive turn-out but it did increase awareness of a candidate.

Sasha Issenberg, author of The Victory Lab, looked at lawn signs for Slate magazine in a 2012 article. Issenberg noted studies that showed a correlation between signs and voting. In one study, “households that displayed either an American flag, football insignia, or campaign sign were 2.4 times more likely to have a resident who voted in the elections than houses which had none of the three.”

-srbp-

Cost per Vote: the Rural/Urban Divide #nlpoli

In looking at by-elections in Newfoundland and Labrador, the second and third phases offer a neat cluster of by-elections in urban (metro St. John’s) and rural (everywhere else) districts.

Break those out and you get some information that shows the relative strengths and weaknesses of the parties, depending on where a by-election occurred during the period between 201 and 2011.

For starters, here is a reminder of the cost per vote results for the three major parties in all three time periods (phases).

Now let’s take a look at the second phase, that is, on the by-elections between the 2003 and 2007 general elections.

Three by-elections took place in what we can consider to be urban.  They are Signal Hill-Quidi Vidi, Kilbride and Ferryland.  The rural by-elections took place in Exploits, Placentia-St. Mary’s, Port au Port, Humber Valley, and Labrador West.

In the urban by-elections, the Conservatives average  cost per vote was $8.98.  The Liberals CPV was $15.71 and the NDP CPV was $18.17.

As much as we all assume that the New Democrats’ base of support is metro St. John’s, these CPV figures suggest otherwise.   Now part of that high cost is attributable to the fact that the Tories hotly contested the seat when Jack Harris left for retirement in 2006. 

But when you average the figure out the average cost per vote of the other two, you get  $15.70. That almost exactly the same as the Liberals and their political infrastructure in the metro St. John’s area just shrivelled up to nothingness by 2007.

CPV is not a measure of actual effort of course.  Parties in Newfoundland and Labrador don’t have to report all their efforts.  Volunteer workers, for example, don’t show up as an expense in and of themselves.  They show up when the party or the campaign has to foot their travel and accommodations bills. 

That’s the kind of thing that happens to the Tories in rural Newfoundland. The Conservative Party itself spent more than $26,000 on campaign worker travel in the Labrador West by-election (2007). The by-elections in Ferryland and Kilbride cost the Tories only about $15,000 and $18,700 for example.

Labrador West essentially comprises two towns that are spitting distance apart.  You would not have to spend $26,000 on travel unless you were flying people in from other parts of the province to work the by-election. Scan the by-election finance reports and you’ll see exactly the same kind of phenomenon in other Tory campaigns in rural Newfoundland. 

Some call it “by-election-in-a-box” but what the Tories really do is maintain a team of campaigners they can drop into augment the local candidate’s effort.  In addition, the Tories have a pattern of tapping into a regular pool of donors to help finance the given by-election.  And again, it is part of a pattern of adding to the local campaign.

Contrast that with the Liberals, for instance.  In any by-election, Liberal candidates are basically on their own.  Sure there are some people who work by-election to by-election, and sure the party headquarters may toss some cash to a campaign. 

But on the whole, Liberal candidates couldn’t count on the party for much of anything during the second and third phase by-elections.  Two noteworthy exceptions to this were the Straits-White Bay North North and Terra Nova both of which took place in 2009 (Phase Three). 

Incidentally, when people talk about the Liberal Party’s lack of organization and infrastructure, this is the kind of stuff they are talking about.  This is the meat and potatoes of politics and the Tories know how to make a fine stew of it.

 

In Phase Three, you had three by-elections in the metro St. John’s are.  Those are our urban set:  Cape St. Francis, Topsail and Conception Bay East-Bell Island.  The rural by-elections were in Baie Verte-Springdale, the Straits-White Bay North,  Terra Nova and Humber West.

The Liberals’ urban cost per vote hit $22.50 in Phase Three. Spending dropped, on average, from about $9600 to a little under $7500 and the average vote went from 612 to $330.

NDP urban spending went from $14,412 to $11, 472 and the vote went from 793 to 751.  The resulting CPV improved in Phase Three, reaching $15.27 compared to $18.17 in Phase Two.

The Conservatives CPV was $9.00.  Their urban spending went up by about $2400 and the average vote went from 2370 (Phase Two) to 2741. 

Conservative rural CPV dropped (marginally) from $23.07 to $22.25.   Tory rural spending dropped by about $6,000 and the average vote dropped by 170 votes.

The New Democrats’ CPV in rural Newfoundland was $30.82 during Phase Three compared to $16.45 in Phase Two. The NDP increased their spending, on average by $1300 but their average vote dropped by 151.  They spent 19% more, in other words, and got 36% less.

Liberal spending on rural by-elections increased by 70%, on average, in Phase Three compared to Phase Two. Liberal rural vote went up by four percent.

- srbp -

Campaign Spending and Efficiency #nlpoli

Two of the three political parties in this province are spending more, on average, and getting less, on average in by-elections.

That’s one of the things you can see in an assessment of a decades worth of by-elections from 2001 to 2011.

Last week, SRBP gave you a teaser of a look at the idea of cost per vote, as measure of campaign efficiency and effectiveness.  As the name implies it tells you how much each campaign spent for each vote it received. The information for the assessment comes entirely from financial reports and by-election vote reports issued by the province’s chief electoral officer.


There were 21 by-elections during that time.  You can break them up into three phases or time periods.  Phase One covers the by-elections between 2001 and the general election in 2003.  Phase Two covers the by-elections from 2004 to 2007.  Phase Three runs from 2008 until the 2011 general election.

The charts below show the average spending and votes received by the three main parties with the resulting cost per vote in each of the three  phases.  We’ll look at the phases individually in other posts.



From Phase One to Phase Two, average Liberal spending on by-elections dropped 64% from an average of around $37,000 to about $13,500. In Phase Three they were spending about $20,000 less per by-election than they were when they were in power.

In the shift from being the opposition to government, Conservative average spending climbed from about $27,000 to $40,700.  It’s a jump of about 51%.

The average NDP spending over the same two periods went up from $2118 to $10,069.

The NDP and the Conservatives spent more in Phase Three than they did in Phase One.  Even the Liberals, who dropped significantly by the second phase had boosted their spending by the third phase to hit $17,663 per by-election on average.

What they got for their efforts is shown in the comparison of average votes received.



The most dramatic changes are at the beginning.  Both the average Tory and average Grit vote per by-election dropped precipitously from Phase One to Phase Two.  The Conservatives went down 30% while the Liberals went down 42%.

Dipper support climbed from 132 votes, on average, to 576.

The Liberal slide continued into Phase Three.  They dropped another 200 votes and took an average 1,000 per by-election in Phase Three.



The line that stands out is the orange one.  The NDP are getting more votes and they are spending more money on average to get them.  But the cost per vote is also climbing.  In the third phase, the NDP was spending $20.23 per vote compared to $16.05 before 2003.

The Conservatives saw a huge increase in their cost per vote by the second phase.  Incumbency effectively doubled their CPV from $8.15 to $17.65.

In Phase Three, the Tory CPV was $15.70.  It’s the lowest of the three parties, but not by much. By Phase Three, the Liberals are back close to their pre-2003 CPV with $17.64.  That’s nothing to cheer about.

Those Phase Three CPV numbers suggest that all three parties have problems getting their messages across or aligning with public opinion.  Remember, the lower the CPV is, the better you are doing.

Still, you can see some indication of what came in the 2011 general election if you look at Phase Three compared to Phase Two:

• The Tories spent 8% less and got 3% more votes. They’ve got a relatively better position than the others and that helps keep them in power.
• The Grits spent 31% more and got 17% less in the vote department.  That’s all sorts of bad news.
• The Dippers spent 5% less and got 18% less in votes.  Again, that should be all sorts of news, most of which isn’t good. 

Once Elections NL releases the 2011 general election financial reports, SRBP can do a comparison for all three general elections since the turn of the century.  We can also cross reference the general elections with the by-elections to see if there are any things that turn up.

In the next post on CPV, we will take a look at the three phases broken down by urban and rural by-elections.

- srbp -

*edited to correct typos

Cost per vote: by-elections #nlpoli

One measure of campaign efficiency and effectiveness is comparing the amount a campaign spends for the number of votes it receives.

It’s called cost-per-vote.

Here’s a comparison of the cost per vote for each of the three political parties in a string of by-elections over the past three years.  From left to right, they are:

• Straits-White Bay North
• Terra Nova
• Topsail
• Conception Bay East-Bell Island
• Humber West

Basically you judge CPV this way: the lower the number, the better. No matter whether your party is the incumbent or one of the outs, you want to spend as little money as possible to get your votes in the box.  That just reflects a basic notion about spending money efficiently.  Money, like people and time, is precious.

Just so that you don’t get turned around here, we are not talking about winning or losing.  Spending the most per vote doesn’t necessarily guarantee success at the polls. In fact, you can spend the most per vote and get clobbered.

In the four by-elections they won, the Tories spent the least amount per vote compared to the other parties. In three of the by-elections, the variation was marginal. In Humber West, for example, the Conservatives spent $21.50 per vote compared to the Liberals $23.48. The New Democrats spent $28.29.

In the one by-election in this series they lost (Straits-White Bay North in 2009), the Conservatives spent $30.96 per vote compared to $19.61 per vote for the Liberals.  The NDP spent $37.50.

Take a look at the three by-elections in between those two extremes and you’ll see other extremes.  The Tory CPV  in Topsail was a mere $6.15.  In Terra Nova it was $11.53.

One of the reasons for this is that the Tories didn’t have to ship in large numbers of campaign workers a long distance from St. John’s in order to effect the win.  In both Humber West and the Straits, the Tories’ centrally controlled campaign system meant they had to pay cash to cover the travel and accommodations of their workers.

They didn’t have to do that in the two by-elections closest to Sin Jawns.  And not surprisingly, that’s the two where the Tory CPV was less than 10 bucks.

As noted here after the Humber West by-election, the Tories poured a heavy effort into hanging onto the seat.  Now that we have actual financial details, we can see that the Tories were relatively more efficient in CPV terms. 

But notice how close together the three parties were.

That’s interesting.

It will get more interesting when you look at the trends in cost-per-vote over the past decade and a bit.  That’s for another post.

- srbp -

Related:

• Buzzfeed:  How much did each Republican candidate spend per vote in Iowa?
• The Atlantic:  the price Rick Perry paid for each vote in Iowa.
• The cost per vote in 2010 Texas state primaries.

JFK’s speech on religion and politics

Religion played a huge role in the 1960 presidential election.

Specifically, Republicans questioned how John Kennedy would govern, given that he was a practicing Roman Catholic.   A group of 150 Protestant ministers and laymen publicly opposed the idea of a Roman Catholic president.  In a public statement, the group  -including Rev. Dr. Norman Vincent Peale – stated that, among other things, they felt it "is inconceivable that a Roman Catholic President would not be under extreme pressure by the hierarchy of his church to accede to its policies with respect to foreign relations.”

While the Republicans did not use the religion issue as part of the national campaign, the issue continued to dog Kennedy throughout the race.

A speech to Greater Houston Ministerial Association in September 1960, right,  proved to be Kennedy’s definitive statement on the subject.

The campaign used film of the speech in television advertising and ran it repeatedly throughout the fall of 1960, especially in areas where there was a heavy Roman catholic population.

In her definitive study of presidential advertising, Kathleen Hall Jamieson demonstrates that the Democrats used Kennedy’s speech both defensively and offensively.  They used it to rebut the Republican attacks and at the same time tried to motivate Roman Catholic voters.

National Public radio produced a transcript of the entire speech.  As Jamieson notes, some consider this to be Kennedy’s best speech. At the opening, Kennedy lists what he considers to be the real issues of the campaign.  

He then turns to a series of statements of his own views, that flow from this introduction:

But because I am a Catholic, and no Catholic has ever been elected president, the real issues in this campaign have been obscured — perhaps deliberately, in some quarters less responsible than this. So it is apparently necessary for me to state once again not what kind of church I believe in — for that should be important only to me — but what kind of America I believe in.

The structure of the speech -  a series of paragraphs starting with “I believe…” deliberately mimics the structure of any statement of faith.  Kennedy continues to recite the articles of his own political faith including the separation of church and state, that effective places his critics in the position of doing the very thing they attack Kennedy on. 

He finishes his statement of political faith with words every member of his audience would know:

But if, on the other hand, I should win the election, then I shall devote every effort of mind and spirit to fulfilling the oath of the presidency — practically identical, I might add, to the oath I have taken for 14 years in the Congress. For without reservation, I can "solemnly swear that I will faithfully execute the office of president of the United States, and will to the best of my ability preserve, protect, and defend the Constitution, so help me God.

You can find the video of the speech on youtube:

JFK on religion and politics

These days, Kennedy’s campaign could not rely on broadcasting the speech in five minute and the full 10 minute airings as part of their ad campaign. 

This is a short speech.

The words are compelling and Kennedy delivers it reasonably well.

The subject was highly controversial in a way that few mi9ght appreciate these days.

But the audience has changed.

Modern audiences simply won’t sit still for a talking head that goes on about any subject for nearly 11 minutes.

They definitely would find Kennedy’s speech itself taxing.  The sentences are much longer than the short statements that modern audiences are used to hearing. While it suits the immediate audience, the speech demands that people be familiar with the subject and with a great deal of history, including more recent events at the time.

Still, there are quotable bits likely crafted to make them fit with a potential series of short spots:

These are the real issues which should decide this campaign. And they are not religious issues — for war and hunger and ignorance and despair know no religious barriers.

The first sentence doesn’t stand on its own, however. 

You’d have to hear the list of real issues to get the full effect.

While Kennedy runs them off effectively enough, the transcript would have great visual impact if the issues came not as a series of clauses separated by semi-colons but as a bulleted list:

I want to emphasize from the outset that we have far more critical issues to face in the 1960 election:

• the spread of Communist influence, until it now festers 90 miles off the coast of Florida;
• the humiliating treatment of our president and vice president by those who no longer respect our power;
• the hungry children I saw in West Virginia;
• the old people who cannot pay their doctor bills;
• the families forced to give up their farms;
• an  America with too many slums, with too few schools, and too late to the moon and outer space.

Still, each one is expressed simply enough.  The last point relies on the repetition of the word “too” – “too many slums”,  “too few schools,” “too late” to space.

All the same, the speech is extraordinarily well written both for its time and for today.

It reflects the input of Kennedy’s political staff – like ted Sorensen – and undoubtedly the campaigns advertising staff as well.

- srbp -

Attack of the Fluffy Bunnies #elxn41

You can tell we are in the final week of a federal election campaign.

The attack ads are out.

Everybody has one.

Some people are whining about them.

Some candidates are trying to distance themselves from them, as in NDP candidate Ryan Cleary:

The NDP is distributing a flyer questioning Ignatieff's trust and attendance in Parliament.

The party's candidate in St. John's South–Mount Pearl, Ryan Cleary, said the negative advertisements are coming from the national NDP campaign, not his.

"I've seen that and that's one of the strategies they want to take, so, yeah," he said.

There’s no small measure of hypocrisy in the NDP attacks, of course.  Cleary’s supporters have been whining about since 2008 about some material Liberal incumbent Siobhan Coady circulated in the last election about Cleary.

There’s a reason why campaigns use so-called attack ads or negative ads:  when they are done properly, they work and they often work better than anything else.

“Done properly” means factual comments that address a concern voters have about a candidate, a party or a leader.

A campaign team will figure out the voter concern through polling.  Forget the horse-race stuff.  That’s just bumpf for the people who need to feed their tweets.   A well-organized campaign can turn around the run-away races or quickly eliminate a close race using solid research coupled with good communications and a willingness to use the tools correctly. 

Make no mistake:  it isn’t that simple or formulaic. Campaigns are as much about the battle between teams as anything else. Teams are made up of people. People make decisions and sometimes they make bad ones.  But given the situation, most campaigns will use the tools they have, including negative or attack ads.

What makes the spate of local attacks ads curious is that they seem to be bizarre.

According to the Telegram, the NDP dropped a postcard in St. John’s East this weekend that, like its counterpart in St. John’s South-Mount Pearl attached Liberal leader Michael Ignatieff because of his poor attendance in parliament?

Okay people.

Jack Harris needs to run attacks ads to fend off the Liberals?  According to the only poll publicly available, the Liberal contender in that riding is running at less than five percent of the vote.  The thing can’t have an spill-over because it was delivered to households.

But then there’s the issue. People loved Danny Williams all to pieces and he had such naked contempt for the legislature he set new record lows for sitting in the legislature by a serving Premier.  On the face of it, this is not what you would call a vote-determining issue.  But hey, maybe they have some sooper sekrit research that shows people in Kilbride and Torbay are parliamentary channel nerds.

Then there are the Conservative ones.  Loyola Sullivan is hammering away at the idea that without a Conservative government in Ottawa the loan guarantee for Muskrat Falls is a goner.

In February, 2011 CRA’s polling for the provincial government showed that three percent of respondents thought that the Lower Churchill should be the top priority of the provincial government in the next few years.

3%.

That’s not a typo.

That’s all.

Even if Loyola was running in Labrador, his Lower Churchill scare piece would appeal to only the 4% of respondents there who thought it was an issue.  Again, not exactly what you would call a topic that is going to drive voters to the polls, let alone make them pick a particular candidate.

For their part the Liberals, their bizarreness is the absence of any local negative pieces this time out. That’s not to say there aren’t;  maybe there are.  It’s just that your humble e-scribbler hasn’t seen or heard anything. 

Not like there isn’t plenty to hammer away at for either the Conservatives or the New Democrats and their candidates just like their could be for blue and orange to toss at red. It’s the kind of stuff you’d expect in a race like St. John’s South-Mount Pearl that is, supposedly, too close to call.

Maybe the clue is the way all the candidates and all the media are fixated on Muskrat Falls. Grits and Dippers are talking about it because the Connies made it an issue. The media are covering it because the pols keep telling them it is big. But there is no sign the voters give a tinker’s damn about the dam.

Bizarro.

No one should be surprised if – in the end - the Newfoundland negative ads in this federal election are likely to fall completely flat.  The reason is simple: they obviously aren’t driven by what voters are interested in. 

As a result, they may look aggressive but they are the equivalent of hitting your opponent with fluffy stuffed bunnies. 

- srbp -

Campaign College: Sign Design 101

There is a difference between political communications and advertising.

Here’s a simple, graphic example. 

One of the signs in these pictures was designed from the start to convey essential information simply, succinctly and at long distances.

It uses bright colours and a font that is easy to read, even at a distance.

It was not designed by an ad agency.

A by-product of the design is that it is also visible – and legible – when travelling at the speeds one normally encounters in city driving.

The other two weren’t.

Obviously.

The pictures were taken at the intersection of Westerland Road and the parkway, from the corner opposite the signs and without any zoom.

The sign legible in this shot was also plainly legible to the unaided eye at twice and three times this distance.

-srbp-