Want a more representative democracy? Reform CT's redistricting process

This opinion was originally published in the CT Mirror.

The Nov. 8, 2022 election was the first based on  the federal and state legislative districts newly drawn  in 2021. Every ten years, in the year following the decennial census, voting districts are redrawn to ensure equal numbers of residents within each district.

The League of Women Voters of Connecticut (LWVCT) wrote extensively about the redistricting process while it was being conducted in 2020 and 2021.

Shortly after the election, The Connecticut Mirror assembled a panel of experts to discuss the results and themes that impacted Connecticut voters. One theme,  the “power of incumbency,” means that once a candidate has been elected their odds of re-election are substantially increased. This phenomenon can be quantified by assessing election results. Of the 151 CT Assembly districts, 126 incumbents were re-elected. Of the 36 Senate districts, 28 incumbents were re-elected. Further, a substantial number of incumbents were uncontested. Of the 151 Assembly seats, 39 were uncontested, per the CT Mirror panel. If races aren’t competitive, candidates have little motivation to participate in debates or provide information for voter guides. This failure to engage with the electorate is a problem throughout the country, not just in CT, as we have previously reported.

This is where the story and process of redistricting comes in. U.S. House representatives and State Assembly members and Senators are not elected “at-large” by voters across the entire state. They are elected from specific districts: These districts can be shaped geographically to favor a political party or an incumbent.

The Connecticut Constitution defines how redistricting is conducted; it specifies a bipartisan multi-step process embedded in the General Assembly.  As we wrote in our blog series, as successive steps are reached and  members are unable to reach agreement, this process becomes increasingly undemocratic.  At stage 1 of the process, a supermajority of Assembly members must agree to the proposed maps prepared by the Reapportionment Committee.  However, in the last 40 years the Reapportionment Committee has been unable to reach agreement; so eight leaders of the General Assembly are appointed to serve on a Reapportionment Commission to broker a deal.

Early in the 2021 redistricting cycle LWVCT learned of a newly available statistical method called “ensemble analysis.” Using this method, analysts can generate thousands of theoretical district maps that adhere to the requirements of population equality and minimization of town border splits. Then we can look to see where incumbent addresses lie in these theoretical, model maps.

Using resources provided by the national League of Women Voters People Powered Fair Maps© campaign, we engaged Kyle Evans of Trinity College to conduct an ensemble analysis of the Connecticut Assembly and Senate approved maps.  In particular we asked Evans to assess the extent to which Connecticut maps appeared to be drawn to favor the incumbent. The key question asked is how likely is it that a map that comports with population equality and minimization of town splits criteria would include the home address of a single incumbent versus multiple incumbents or no incumbents.

As the analysis shows, about 52% of the statistically-generated or “model” maps included a single incumbent. This means 47.8% of the model House maps included either no incumbent or two or more incumbents. In contrast, about 97% of the maps adopted by the bipartisan commission had a single incumbent. These are the districts that were used in the November 2022 election and will be used for the next ten years.

As Evans and his colleague summarized: “These results indicate the 2021 CT State House map is an extreme outlier in terms of incumbent placement in newly drawn district boundaries.”  That is, these one-incumbent-to-a-district maps were unlikely to have been chosen by chance.  Similar results are found in the analysis of State Senate districts.

This analysis suggests that the adopted district maps are indeed structurally defined in a manner that advantages incumbents. Thus the election results described earlier are seen to be in part a rather predictable outcome of the underlying structures (maps) created by the General Assembly. This does not mean that the maps drawn to favor incumbents are the sole determinant of the election results; certainly the candidates and campaigns themselves are relevant inputs to the outcome!

In the most recent election there is the example of Assembly District 81 where the winner prevailed by one vote.  Notably, however, this was a district in which the incumbent retired, meaning neither candidate benefited from an incumbent advantage.

Also, it can be correctly noted that incumbency can bring advantages to constituents. One such advantage is institutional knowledge and power. On the other hand, a widely acknowledged byproduct of the incumbency effect is that it constitutes a barrier to new entrants to the field of representative government with the result that political institutions are populated with older, whiter, and many more male representatives than the population from which these advantaged candidates are drawn.

In other words, a redistricting process that favors incumbents tends to ossify our political establishment. Indeed, the incumbency advantage/barrier is so well known that programs training women as potential candidates advise them to move to a more favorable district to increase their chances of being elected.

By this point, readers may well be asking “are there no remedies to this situation?”  Throughout the past redistricting cycle, LWVCT advocated for several reforms to strengthen public involvement in the redistricting process. Some of our espoused reforms could be implemented without amending the state Constitution. 

Beyond reform, the League’s national position favors removing redistricting from elected bodies and establishing redistricting commissions independent of seated representatives.  We are convinced this is the approach that would be best for Connecticut.

You can read more about redistricting commissions, including independent redistricting commissions, using these resources: Redistricting commissions – Ballotpedia ; On Movement to Establish Citizen-Led Redistricting Commissions in the States or read this article as a cautionary tale describing commission failure and identifying necessary commission elements to avoid gridlock.

In summary the LWVCT recommends:

1. Use modern tools such as ensemble analysis to guide districting decisions;
2. Remove the veil shielding the redistricting process from public view. Currently redistricting is accomplished behind closed doors. Greater transparency is needed all along the way including but not limited to establishing a delineated set of steps prior to adopting maps. For example, some states “show their work” by publishing draft maps and soliciting public input before final acceptance.
3. Educate and inform the public about redistricting alternatives, especially independent redistricting commissions.

It is our hope that by pointing out that the current process is shielded from view and results in overprotection of incumbents, that when the next census occurs ten years from now, Connecticut will be ready to implement a more participatory system of redistricting. It is surely time for Connecticut to live up to its progressive credentials and create maps that represent communities and encourage new entrants.

 Joan Twiggs is Redistricting Champion and Laura Smits, President, respectively, of the League of Women Voters of Connecticut.