Posts tagged INST316

Teaching About Electoral Systems

This semester I had a chance to develop and teach a seminar on electoral systems and party systems. The course has been heavily metrics based, with a strong emphasis on how to calculate a number of metrics used in the study of elections and parties. 

So far so good, with most of the students successfully learning how to calculate the effective number of parties and electoral volatility. They were tested on this on the midterm: I was pleased that, given a table of data (for a fictitious country, Oz), they were able to calculate both measures in under an hour (while also working on a short essay question).

For the final exam, I’m upping the ante: Given a table of election results, they’ll have to calculate the seat distribution using a “largest remainder" system (Imperiali) and a "highest average" system (pure Sainte-Laguë). They’ll then have to also decide wether a 7% electoral threshold would alter the seat distribution in any way.

adam-wola:

“Shock,” Ana Tijoux’s tribute to Chile’s burgeoning student protest movement. See also the video for “Sacar la Voz.”

I was actually going to use this in class today while we discussed Chile’s unique binominal electoral system.

The unique electoral system—which uses two-seat districts with an open-list PR formula—is a legacy of the Pinochet era. It was consciously designed to ensure that conservatives (about 1/3 of the electorate) could ensure nearly 50% of the seats.

Two decades after the transition to democracy, many Chileans (particularly younger ones with little or no memory of the Pinochet dictatorship) are demanding reforms to the Pinochet-imposed institutional framework.

BTW, you should check out Ana Tijoux’s autobiographical “1977”.

Talking about the single transferable vote (STV) electoral system in INST 316 (Electoral Systems & Party Systems). STV is used in only a small handful of countries in the world—one of which is Ireland. We went over how it works on Monday, took a break to talk about linear regression on Wednesday, and today will have a student report on how STV affects Irish elections.

I thought it might be useful to help explain the Irish political system—and particularly its deep roots in the Irish Civil War of 1922-23—through music. The war came on the heels of Irish independence, which began in earnest with the Easter Rising of 1916 (which inspired the famous ballad, “Foggy Dew”). If you’re looking for a great Irish history site, there’s none better than the National Library of Ireland, which has a great online exhibit of the 1916 Rising.

So I’m playing for my students the Black 47 song “Big Fellah.” I think the song does a good job of laying the groundwork for the basic framework for understanding post-independence Irish politics: The Irish rose up in 1916 using the [Dublin] GPO (General Post Office) as headquarters, fought together against the “black & tans” (the English), but then became divided over the terms of the peace treaty (which “came back with an Ireland divided up in two!”), then fought a civil war amongst themselves. The song also focuses on Michael Collins (the subject of a recent film, staring Liam Neeson), who was later killed by his former partisans because he backed the partition settlement.

If you want to watch a more powerful film on the subject, I highly recommend The Wind that Shakes the Barley. It’s the tragic story of two brothers who fight for Irish independence, then end up on different sides in the Irish Civil War. That film, perhaps more than any other, shows the political divisions between what later became the two most powerful parties in Ireland: Fianna Fáil (the party that opposed the partition treaty) and Fine Gael (the party that supported the partition treaty).

Enough of the students in INST 316 are doing well, that I’m going to go ahead and teach them the new Golosov (2010) formula for effective number of parties. I’ve started using it myself because it provides more information than the traditional Laakso & Taagepera formula.
Not only is the Golosov formula better at discriminating multiparty systems with dominant parties from those that are hyper-competitive (lots of little parties), it provides some nice cross-party information as well. Unlike the Laakso & Taagepera formula, it can be computed in two steps. Doing so, gives a relative “size” to each individual party (pi). Automatically, the largest party is “1” and all others are relative to its size (for example, a party can be “0.237” or about a quarter of a party). Since all you do is add up the bits of parties to come up with the total, you can nicely compare the relative differences in size between parties—or the relative balance of government/opposition blocks.
—-
Golosv’s formula first appeared in the March 2010 issue of Party Politics. However, I highly recommend his follow up in the September 2011 issue of Party Politics, in which he defends the the formula and demonstrates its use for distinguishing different kinds of party systems using a Nagayama diagram (basically a geometric figure using conceptually defined data points to create spaces to locate observed data within a typology).

Enough of the students in INST 316 are doing well, that I’m going to go ahead and teach them the new Golosov (2010) formula for effective number of parties. I’ve started using it myself because it provides more information than the traditional Laakso & Taagepera formula.

Not only is the Golosov formula better at discriminating multiparty systems with dominant parties from those that are hyper-competitive (lots of little parties), it provides some nice cross-party information as well. Unlike the Laakso & Taagepera formula, it can be computed in two steps. Doing so, gives a relative “size” to each individual party (pi). Automatically, the largest party is “1” and all others are relative to its size (for example, a party can be “0.237” or about a quarter of a party). Since all you do is add up the bits of parties to come up with the total, you can nicely compare the relative differences in size between parties—or the relative balance of government/opposition blocks.

—-

Golosv’s formula first appeared in the March 2010 issue of Party Politics. However, I highly recommend his follow up in the September 2011 issue of Party Politics, in which he defends the the formula and demonstrates its use for distinguishing different kinds of party systems using a Nagayama diagram (basically a geometric figure using conceptually defined data points to create spaces to locate observed data within a typology).

This is what I’m teaching tomorrow in INST 316. We’ll be using seat volatility, rather than electoral volatility. But it’s our second follow up to last week’s workshop on calculating the effective number of parties.
This semester I’m trying an interesting approach. My course on Electoral Systems & Party Systems is a methods course in disguise. Each student has been assigned a country, and has been responsible to collect party seat data for that country since 1990. After building their own small dataset, they’re going to calculate some metrics (with a focus on effective number of parties and volatility—but the textbook introduces a range of others). Later, we’ll do some simple bivariate linear regression (using Excel, and despite the limitations of a very small dataset) and, much later, once we combine all the datasets, we’ll test for relationships between party system indicators and type of electoral system with some simple Chi-square tests (again, using Excel).
So we’re not going to cover some of the quantitative tools that are better done with a dedicated statistics package (like Stata or SPSS), but at least they’ll have some basic framework for when they do take a methods course later (or to reinforce their previous methods course).
However, students have the chance to write short reports on each of the two metrics (they already did a great job on the effective number of parties assignment). The papers are very simple, three-page reports that also include a table (presenting the data) and a chart (showing the trend in the index over time). One of the things this has emphasized for me is that students do not really know how to use Word or Excel effectively. 
All the while, I hope they gain a better appreciation for our discussion about party systems and electoral systems in various countries by really looking at how the conceptual debates are rooted in empirical analysis.

This is what I’m teaching tomorrow in INST 316. We’ll be using seat volatility, rather than electoral volatility. But it’s our second follow up to last week’s workshop on calculating the effective number of parties.

This semester I’m trying an interesting approach. My course on Electoral Systems & Party Systems is a methods course in disguise. Each student has been assigned a country, and has been responsible to collect party seat data for that country since 1990. After building their own small dataset, they’re going to calculate some metrics (with a focus on effective number of parties and volatility—but the textbook introduces a range of others). Later, we’ll do some simple bivariate linear regression (using Excel, and despite the limitations of a very small dataset) and, much later, once we combine all the datasets, we’ll test for relationships between party system indicators and type of electoral system with some simple Chi-square tests (again, using Excel).

So we’re not going to cover some of the quantitative tools that are better done with a dedicated statistics package (like Stata or SPSS), but at least they’ll have some basic framework for when they do take a methods course later (or to reinforce their previous methods course).

However, students have the chance to write short reports on each of the two metrics (they already did a great job on the effective number of parties assignment). The papers are very simple, three-page reports that also include a table (presenting the data) and a chart (showing the trend in the index over time). One of the things this has emphasized for me is that students do not really know how to use Word or Excel effectively. 

All the while, I hope they gain a better appreciation for our discussion about party systems and electoral systems in various countries by really looking at how the conceptual debates are rooted in empirical analysis.

Why Tuesday? | The Monkey Cage

From The Monkey Cage:

A reader points me to the group, Why Tuesday, that wants to move Election Day to a more convenient time.  They write:

Today, we are an urban society, and we all know how hard it is to commute to our jobs, take care of the children, and get our work done, let alone stand on lines to vote. Indeed, Census data over the last decade clearly indicates that the inconvenience of voting is the primary reason Americans are not participating in our elections.

If we can move Columbus Day, Presidents’ Day, and Martin Luther King, Jr.’s Holiday for the convenience of shoppers, why not make Election Day more convenient for the sake of voters? First and foremost, it is time to end the deafening silence of good people on this vitally important issue. So we ask: Why Tuesday?

Personally, I would have no problem with this.  But I’m not sure it’s going to increase turnout.  The political scientist André Blais reviewed a lot of evidence on turnout for a chapter in this book.  Here is what he wrote on Sunday voting and “rest day” voting (links added by me):
Unfortunately, we know relatively little about the actual impact of such measures on turnout. Franklin (1996) initially reported that postal (absentee) voting and Sunday voting increase turnout but his subsequent analysis of turnout changes (Franklin 2004) indicate no independent impact.  Norris (2004) examines the effect of a variety of rules (number of polling days, polling on rest day, postal voting, proxy voting, special polling booths, transfer voting, and advance voting) and finds no effect….It makes sense to assume that people are more willing to vote when it is easy than when it is difficult.  But we still do not know which measures are the most efficient or how much difference they make.

Again, I’d certainly be amendable amenable to seeing if non-Tuesday voting made any difference.  I’m just cautious in believing it would.

The only formula I know that independently increases turnout, is making it mandatory. Voting is mandatory in most countries in Western Europe and Latin America. And even though enforcement is weak (if at all), such countries have significantly higher turnout (roughly 70-80% in Latin America and around 90% in some West European countries).

Here’s an excellent article on voter turnout in comparative perspective, from International IDEA.

This is what I’m teaching today in INST 316. Yes, I know that Golosov’s (2010) equation is better.

This is what I’m teaching today in INST 316. Yes, I know that Golosov’s (2010) equation is better.

Piñera and [Chile's] binomial system

Greg Weeks at Two Weeks Notice comments on the possibility that Chile’s president, Sebastian Piñera, may be joining the (growing?) bandwagon for electoral system reform in Chile.

It’s only inevitable, I suppose, that electoral system reform fever would reach Chile. I’m partly sad, because Chile’s binomial system—a platypus of an electoral system made for interesting teaching (as did Uruguay’s ley de lemas, or double simultaneous voting, system).

Obviously, the binomial system is just a holdover from the Pinochet era, when the outgoing dictatorship sought to safeguard its policies and virtually ensure a (slim) conservative majority. That system has outlived its use, and electoral volatility and voter antipathy is already eroding the party system tied to the binomial system, making it increasingly unwieldy. But what would replace it? And what would its effects be on the party system—and on governance—in Chile?

(I ask these questions with an eye to my INST 316 seminar on electoral systems and party systems, in which Chile is one of our case studies.)

INST 316: Electoral Systems & Party Systems

If you’re interested in electoral systems & party systems, consider taking my Croft seminar next semester. Here’s a link to the syllabus. Space is limited (to 15 students), so enroll soon!

Course Description:

This is a seminar on electoral systems and party systems, and their relationship to “governance” across different democracies. While democracy is now common in many regions of the world (particularly Europe and Latin America), not all democracies have similar institutions. This course focuses on two key institutions—electoral systems and party systems—that are closely interrelated. Electoral systems are the formal rules by which voters choose their representatives. These formal rules affect how voter preferences translate into seats.

Party systems are the constellation of individual political parties in a country. The structure and historical evolution of party systems—as well as the strategies and trajectories of individual parties—are shaped by (along with other factors) the electoral rules. Lastly, these two variables also affect the quality of “governance” (the way political authority is exercised) in different countries. We will explore the relationship between all three during the semester.

Although the course covers a comprehensive range of electoral systems and party systems, we will focus particular attention on a few key countries in the four Croft area concentrations:

  • Latin America: Brazil, Chile

  • Europe: Britain, France, Germany, Ireland, Sweden

  • Asia: Japan, South Korea

  • Middle East: Turkey

Each student will also have an opportunity to work throughout the semester on a research portfolio exploring a country of his/her choice (in consultation with the instructor). If we have a larger enrollment, we will add one or more of the following cases: Costa Rica, Poland, Spain, Thailand, and Israel. 

INST 316: Electoral Systems & Party Systems

If you’re interested in taking a seminar on electoral systems and party systems next semester (Spring 2012) especially designed for students at the Croft Institute for International Studies at the University of Mississippi:

Course Description:

This is a seminar on two interrelated institutions of democratic politics: electoral system and party systems. The first half of the semester explores the different electoral systems used in selected democracies: France, Germany, Britain, Ireland, Sweden, Japan, South Korea, Turkey, Brazil, and Chile. The second half of the semester explores how party systems vary across countries and how these differences affect country “governance.”

Course Goals:

The course will familiarize students with the politics of different democracies around the world. Additionally, students will become familiar with both qualitative and quantitative analytical frameworks, methods, and metrics used to study political institutions. Conscious effort will made to helping students apply these tools to future theses projects.

Course Texts:

David M. Farrell. Electoral Systems: A Comparative Introduction. London: Palgrave, 2011.

Kenneth Janda. Party Systems and Country Governance. Boulder: Paradigm Publishers, 2011.