Loading...
HomeMy WebLinkAbout04-02-26 Study Commission Agenda and Packet MaterialsA. Call to Order with Pledge of Allegiance and Moment of Silence - 4:00 PM, Commission Room, City Hall, 121 North Rouse B. Changes to the Agenda C. Public Comment on Anything within the Jurisdiction of the Study Commission THE STUDY COMMISSION OF BOZEMAN, MONTANA SC AGENDA Thursday, April 2, 2026 How to Participate: If you are interested in commenting in writing on items on the agenda please send an email to govreview@bozeman.net prior to 12:00 p.m. on the day of the meeting. At the direction of the Study Commission, anonymous public comments are not distributed to the Study Commission. Public comments will also be accepted in-person and through video conference during the appropriate agenda items but you may only comment once per item. As always, the meeting will be recorded and streamed through the Meeting Videos and available in the City on cable channel 190. For more information please contact Ex Officio, Mike Maas, 406.582.2321, or visit bozemanstudy.com. This meeting will be held both in-person and also using an online video conferencing system. You can join this meeting: Via Video Conference: Click the Register link, enter the required information, and click submit. Click Join Now to enter the meeting. Via Phone: This is for listening only if you cannot watch the stream, channel 190, or attend in- person United States Toll +1 669 900 9128 Access code: 951 6442 0347 This is the time to comment on any matter falling within the scope of the Bozeman Study Commission. There will also be time in conjunction with each agenda item for public comment relating to that item but you may only speak once per topic. Please note, the Study Commission cannot take action on any item which does not appear on the agenda. All persons addressing the Study Commission shall speak in a civil and courteous manner and members of the audience shall be respectful of others. Please state your name, and state whether you are a resident of the city or a property owner within the city in an audible tone of voice for the record and limit your 1 D. Consent Agenda D.1 Study Commission Claims Review and Approval(Heinen) D.2 Approval of Study Commission Minutes(Heinen) E. Correspondence or Study Commission Update E.1 Final Survey Results by Working Ventures (Franks/Working Ventures) F. Unfinished Business F.1 Further Learning for Wards/Districts(Clark ) G. New Business G.1 Discussion with Jeff Krauss and Danielle Rogers on City Boards G.2 Learning session for Neighborhood Associations and Boards (Clark) H. Future Agenda Items H.1 Future Education Discussion and Needs for April 15 Meeting (Heinen) I. Public Comment on Anything within the Jurisdiction of the Study Commission J. Announcements K. Adjournment comments to three minutes. Written comments can be located in the Public Comment Repository. Outstanding Invoices for Working Ventures, Bozeman Daily Chronicle, and the MSU Exponent Consider the Motion: I move to approve the Study Commission meeting minutes from March 25, 2026. Reference Materials Study Commission Bylaws Study Commission Resources Study Commission meetings are open to all members of the public. If you have a disability that requires assistance, please contact the City of Bozeman's ADA Coordinator, David Arnado, at 406.582.3232. Study Commission meetings are televised live on cable channel 190 and streamed live on our Meeting Videos Page. 2 Memorandum REPORT TO:Study Commission FROM:Caeleb Heinen, Recording Secretary Mike Maas, Ex Officio SUBJECT:Study Commission Claims Review and Approval MEETING DATE:April 2, 2026 AGENDA ITEM TYPE:Finance RECOMMENDATION:Outstanding Invoices for Working Ventures, Bozeman Daily Chronicle, and the MSU Exponent STRATEGIC PLAN:7.5. Funding and Delivery of City Services: Use equitable and sustainable sources of funding for appropriate City services, and deliver them in a lean and efficient manner. BACKGROUND:Study Commission claims for approval prior to payment by Ex Officio from the approved Study Commission Budget. Budget Tracking UNRESOLVED ISSUES:None ALTERNATIVES:The Study Commission could decide not to approve these claims or a portion of the claims presented. This alternative is not recommended as it may result in unbudgeted late fees assessed. FISCAL EFFECTS:$14,701.90 Attachments: Exponent-Invoice-00025-_-BCSC-_-02-26-2026.pdf BCSC Invoice #7 (March 2026) - Invoice.pdf BCSC Invoice #4 (December 2025) - Invoice (1).pdf BCSC Invoice #6 (February 2026) - Invoice.pdf 75385_0.pdf Report compiled on: March 25, 2026 3 February 26, 2026BCSC Invoice #00025 Contact Name Ana Lopez Shalla Email ana@working-ventures.com Phone Number Run Start Date Thursday, February 5, 2026 End Date Thursday, April 30, 2026 Notes: Discount Applied: Coupon code - 15off$363.00 Discount $363.00 Total $2,057.00 Description Amount Half Page Color $187.00 x 11 $2,057.00 Payment Terms: · Payment is due within 30 days of the invoice date. · A late fee of 5% per month will be added to all past-due balances. · Please include the invoice number with your payment. Please submit payment to: MSU Departments : MSU Exponent Index #431316 - 53350 PO BOX 174200 Bozeman, MT 59717 We appreciate your support of student journalism!4 Working Ventures 2215 S 86th Street Omaha, NE, 68124 EIN: 99-1074520 Invoice Submitted on 03.24.2026 Invoice for Payable to Invoice # Bozeman City Study Commission Working Ventures LLC 26-1004 Project Due date Communications Strategist 4/24/2026 Description Qty Unit price Total price Study Commission Communications Strategist - February 2026 1 $3,900.00 $3,900.00 Amount previously invoiced $27,300 $0.00 *NOTE - Payment for invoice 26-1001 has not been received $0.00 $0.00 Notes: Invoice 1 of 1 Subtotal $3,900.00 Services include: contractual related work for March 2026 Comm Strategist Adjustments $0.00 Note - Final invoice based on early completion of the project.$3,900.00 Invoices will be submitted monthly in the amount of $3,900 per month or 1/12 the cost of the overall project ($46,800) 5 Working Ventures 2215 S 86th Street Omaha, NE, 68124 EIN: 99-1074520 Invoice Submitted on 01.03.2026 Invoice for Payable to Invoice # Bozeman City Study Commission Working Ventures LLC 26-1001 Project Due date Communications Strategist 2/3/2026 Description Qty Unit price Total price Study Commission Communications Strategist - December 2025 1 $3,900.00 $3,900.00 Amount previously invoiced $15,600 $0.00 Amount previously received $11,700 $0.00 Amount yet to be invoiced $27,300 $0.00 Notes: Invoice 1 of 1 Subtotal $3,900.00 Services include: contractual related work for December 2025 Comm Strategist Adjustments $0.00 Note Previous invoice was submitted for Aug-Nov 2025 $3,900.00 Invoices will be submitted monthly in the amount of $3,900 per month or 1/12 the cost of the overall project ($46,800) 6 Working Ventures 2215 S 86th Street Omaha, NE, 68124 EIN: 99-1074520 Invoice Submitted on 03.03.2026 Invoice for Payable to Invoice # Bozeman City Study Commission Working Ventures LLC 26-1003 Project Due date Communications Strategist 4/3/2026 Description Qty Unit price Total price Study Commission Communications Strategist - February 2026 1 $3,900.00 $3,900.00 Amount previously invoiced $23,400 $0.00 $0.00 $0.00 Notes: Invoice 1 of 1 Subtotal $3,900.00 Services include: contractual related work for February 2026 Comm Strategist Adjustments $0.00 Note Previous invoice was submitted for 08/25-01/26 $3,900.00 Invoices will be submitted monthly in the amount of $3,900 per month or 1/12 the cost of the overall project ($46,800) 7 BILLED ACCOUNT NAME AND ADDRESS Pub Description Units ADVERTISING INVOICE /STATEMENT Dates CITY OF BOZEMAN CITY HALL TAKAMI CLARK PO BOX 1230 BOZEMAN MT 59771 2/28/20261 BILLING PERIOD 2/1/2026 - 2/28/2026 BILLED ACCT# 75385 ADVERTISER/CLIENT NAME City of Bozeman City Hall TOTAL AMOUNT DUE $1,432.35 AMOUNT PAID $0.00 TERMS OF PAYMENT Net 30 Days CURRENT NET AMT $944.90 60 DAYS $0.00 90+ DAYS $0.00 30 DAYS $487.45 Net AmountGross AmountAD# BILLING DATEPAGE #REMITTANCE ADDRESS PLEASE DETACH AND RETURN UPPER PORTION WITH YOUR REMITTANCE TYPE COMM PO Num INVOICE #022675385 PO Box 1190 2820 W College, 59718 Bozeman, MT 59771 C/O ISJ Payment Processing PO Box 1570 Big Sky Publishing Pocatello, ID 83204 Big Sky Publishing 406-582-2637 BALANCE FORWARD 487.45 708873 15.00 944.90BDC Know Your City Charter Ad 944.9002/04/26 BDC 708885 15.00 0.00BDC Added Value Know Your City 0.0002/14/26 BDC $1,432.35 TOTAL AMOUNT DUECURRENT AMT DUE $487.45 30 DAYS $0.00 60 DAYS $0.00 90+ DAYS ADVERTISER/CLIENT NAME City of Bozeman City Hall BILLING PERIOD 2/1/2026 - 2/28/2026 ADVERTISER/CLIENT NUMBER 75385 ADVERTISER INFORMATION $944.90 INVOICE AGING OF PAST DUE AMOUNTS *UNAPPLIED AMOUNT *UNAPPLIED AMOUNTS ARE INCLUDED IN NET AMOUNT DUE ADAM'S PUBLISHING RESERVES THE RIGHT TO CHARGE A 1.5% PER MONTH LATE FEE ON ALL AMOUNTS NOT PAID IN FULL WITHIN 60 DAYS. NO ACTION OR OMISSION BY ADAM'S PUBLISHING COMPANY SHALL BE DEEMED A WAVIER OF ITS RIGHT TO CHARGE THIS FEE. $0.00 Invoice # 022675385 C/O ISJ Payment Processing, PO Box 1570, Pocatello, ID 83204 406-582-2637 8 Memorandum REPORT TO:Study Commission FROM:Caeleb Heinen, Recording Secretary Mike Maas, Ex Officio SUBJECT:Approval of Study Commission Minutes MEETING DATE:April 2, 2026 AGENDA ITEM TYPE:Citizen Advisory Board/Commission RECOMMENDATION:Consider the Motion: I move to approve the Study Commission meeting minutes from March 25, 2026. STRATEGIC PLAN:1.1 Outreach: Continue to strengthen and innovate in how we deliver information to the community and our partners. BACKGROUND:Attached are the written minute summaries from the previous two meetings. Future meeting minutes ought to be approved at the next schedule Study Commission meeting. All past meeting recordings are available for review on the City's Meeting Videos page. UNRESOLVED ISSUES:None identified ALTERNATIVES:As per the Study Commission FISCAL EFFECTS:None Attachments: 03-25-26 Study Commission Meeting Minutes.pdf Report compiled on: March 25, 2026 9 Bozeman City Commission Meeting Minutes, March 25, 2026 Page 1 of 3 THE STUDY COMMMISSION MEETING OF BOZEMAN, MONTANA MINUTES March 25, 2026 Present: Carson Taylor, Becky Franks, Barb Cestero, Deanna Campbell, Jan Strout, Mike Maas Absent: None Excused: None A) 00:01:46 Call to Order with Pledge of Allegiance and Moment of Silence - 4:00 PM, Commission Room, City Hall, 121 North Rouse B) 00:03:09 Changes to the Agenda 00:03:21 Cmr. Chair Taylor suggested moving F.1 to follow New Business C) 00:03:43 Public Comment on Anything within the Jurisdiction of the Study Commission 00:04:07 Dan Kaveney gave public comment 00:08:18 Cmr. Taylor responded to public comment D) 00:10:11 Consent Agenda D.1 00:10:17 Approval of Study Commission Minutes 03-05-26 Study Commission Meeting Minutes.docx 00:10:26 Motion to approve I move we approve the Consent Agenda for March 25th. Barb Cestero: Motion Deanna Campbell: 2nd 00:10:40 Vote on the Motion to approve I move we approve the Consent Agenda for March 25th. The Motion carried 5 - 0. Approve: 10 Bozeman City Commission Meeting Minutes, March 25, 2026 Page 2 of 3 Carson Taylor Becky Franks Barb Cestero Deanna Campbell Jan Strout Disapprove: None E) 00:11:19 Correspondence or Study Commission Update E.1 00:11:29 Final Survey Report for Community Engagement 2026.03.16 - BCSC - March '26 Communications Strategy Report.pdf E.2 00:11:34 New Calendar plan 00:11:45 Cmr. Cestero introduced the calendar plan 00:14:13 Cmr. Strout discussed the meeting with League of Women Voters with Cmr. Franks, and the upcoming outreach with Business and Professional Women and Pecha Kucha (September). 00:15:07 Cmr. Taylor discussed a guest editorial in the Bozeman Chronicle by Cmrs. Campbell and Cestero 00:15:33 Ex Officio Maas discussed the upcoming RFP for Ballot Education and the process for selection 00:18:45 Cmr. Campbell requested City Attorney Greg Sullivan to weigh in on the RFP process G) 00:19:09 New Business G.1 00:19:17 Learning session for Wards/Districts At-large VS Wards 3.12.26.pdf Wards Documents for Bozeman 2026.pdf 1. district-based-elections-and-class-based-representation-evidence-from-the-california- voting-rights-act (1).pdf 2. pacificresearch.org-Free Cities Center Analysis City representation Single-member districts versus at large (1) (1).pdf 3. MIT At Large Elections and Minority Representation in Local Government.pdf 4. Primer_DistrictvsAt-LargeElections_Digital-1.pdf 5. Trade-Offs between at-large and single member districts-UofHouston.pdf 6. Wards at-large in Canadian Cities Candian Journal of Poli Sci.pdf 7. Wards vs At large on Diversity.pdf 8. At-Large versus Ward -Implications for Public Infrastructure.pdf 9. Hybrid City Council Election Systems.pdf City Comparison Chart 00:19:27 Dan Clark, MSU Local Government Center, began the discussion on Wards and Districts 11 Bozeman City Commission Meeting Minutes, March 25, 2026 Page 3 of 3 01:24:03 Public comment on wards/districts 01:24:15 Mary Bateson gave public comment 01:28:36 Terry Quatraro gave public comment 01:32:16 Natsuki Nakamura gave public comment F) 01:36:18 Unfinished Business F.1 01:36:24 Review and Discuss Suggested Edits for Article III Draft Ariticle III DRAFT Suggested Edits.docx.pdf Bozeman City Charter 01:36:41 Cmr. Cestero introduced the suggested edits for Article III in draft Charter 01:37:57 Comments from Cmrs. on the suggested edits for Article III H) 01:40:41 Future Agenda Items H.1 01:40:52 Future Education Discussion and Needs for April 2 and April 15 Meetings 01:41:06 Discussion on upcoming April meetings including guest speakers and extra data I) 01:53:29 Public Comment on Anything within the Jurisdiction of the Study Commission 01:53:36 Daniel Carty gave public comment 01:54:31 Mary Bateson gave public comment 01:55:55 Gracie Caldwell gave public comment J) Announcements K) 01:57:34 Adjournment 12 Memorandum REPORT TO:Study Commission FROM:Caeleb Heinen, Recording Secretary Mike Maas, Ex Officio Becky Franks, Study Commission Vice Chair Working Ventures SUBJECT:Final Survey Results by Working Ventures MEETING DATE: April 2, 2026 AGENDA ITEM TYPE: Citizen Advisory Board/Commission RECOMMENDATION: View and discuss the Final Survey Report created by Working Ventures for the four complete surveys and their analysis. STRATEGIC PLAN: 1.1 Outreach: Continue to strengthen and innovate in how we deliver information to the community and our partners. BACKGROUND: A Final Report on the Surveys conducted by Working Ventures in collaboration with the Bozeman Study Commissioners. UNRESOLVED ISSUES: None identified ALTERNATIVES: As per the Study Commission FISCAL EFFECTS: None identified Attachments: Survey Results Final.pdf Report compiled on: March 25, 2026 13 1 LIFTING ALL VOICES FOR LOCAL GOVERNMENT Final Survey Results March 2026 City of Bozeman, MT 14 2 Table of Contents Four survey were completed during this project. Community Survey - (October - November 2025) City Board Survey (January - February 2026) Employee Survey (January 2026) EACH SURVEY WAS DEVELOPED IN COLLABORATION WITH THE BOZEMAN CITY STUDY COMMISSIONERS City Commissioner Survey ( February - March 2026) OverviewKey Components SURVEYS Community Survey Employee Survey City Board Survey Commissioner Survey 3 6 11 17 15 Strongly Disagree Strongly Agree Average 18-64 years 65+ years 1 2 3 4 5 about local government issues in Bozeman. who leads and makes decisions for our city. nt structure of city government in Bozeman. I stay informed about localgovernment issues in Bozeman. I understand who leads andmakes decisions for our city. I am confident in the currentstructure of city government inBozeman. 3 CURRENT STATE PERCEPTIONS RESPONDENTS FEEL INFORMED ABOUT THE CITY, BUT HAVE LOWER CONFIDENCE AND MIXED PERCPETIONS OF STRUCTURE Current State Word Cloud Survey respondents provided the topwords they would use to describe thecurrent structure of local government. Understanding & AttitudesSurvey respondents rated items on a scale from 1 (strongly disagree) to 5(strongly agree). Working age (18-64; n=299) respondents generally had morepositive perceptions than did older (65+; n=163) respondents. LEADERSHIP PRIORITIES TOP PRIORITIES INCLUDE ACCOUNTABILITY, TRANSPARENCY, RESPONSIVENESS, FISCAL RESPONSIBILITY & DEMOCRACY 0 50 100 150 200 250 300 Accountable Leadership Transparent Leadership Responsive Administration Fiscally-Responsible Model of Government Democratic Representation Efficient Administration Stable Administration Term-Limited Leaders Visible Leaders Other Leadership PrioritiesSurvey respondents voted on theirtop 3 priorities for how they wantthe City of Bozeman leadership tofunction going forward. Community Survey 16 4 ELECT ADDITIONAL OFFICIALS RESPONDENTS SHOW SOME INTEREST IN IDEA OF ELECTING ADDITIONAL OFFICIALS, BUT MORE INFORMATION IS REQUIRED CITY MANAGER ROLE Survey respondents selected whether they would support theidea of having additional city officials elected by voters. Forexample, in some cities, positions such as City Clerk, CityTreasurer, or City Attorney are elected. Yes, I would support electing more city officials40.3% No, I would prefer these positions remain appointed 34.4% Not sure / Need more information 25.3% CITY COMMISSION FEEDBACK RESPONDENTS SHOW SOME PREFERENCE FOR GEOGRAPHIC REPRESENTATION BUT MIXED OPINIONS ON EXPANSION GEOGRAPHIC VS. CITY-WIDE COMMISSIONERSSurvey respondents rated their preference on a scalefrom 1 (strongly disagree) to 5 (strongly agree). I believe Bozeman should prioritize geographic district- based representation rather than total city-wide votes for commissioners. Strongly Disagree Strongly Agree Overall Quadrant OutsideOverallQuadrantOutside 1 2 3 4 5 Overall (n=523) NW (n=168) Outside City Limits (n=12) NE (n=119) SW (n=36) SE (n=182) 3.4 3.6 3.6 3.3 3.3 3.2 POTENTIAL GEOGRAPHIC RULESSurvey respondents also considered IF geographic-basedcommissioners were used, which rules would make the most sense. 0 50 100 150 200 250 300 Commissioners live in district, Voters only from district Commissioners live in district, Voters from across the city Commissioners live across city, Voters only from district No preference COMMISSION EXPANSIONSurvey respondents also selected whetherthey thought Bozeman should expand itsCity Commission beyond 5 members. No41.7% Yes33.7% Unsure24.6%Public stakeholdermeetings generatedthe idea of a hybrid model with bothward-based and at-large commissioner elections Community Survey 17 5 ADVISORY BOARDS & NEIGHBORHOOD COUNCILS FEEDBACK ADVISORY BOARDS & NEIGHBORHOOD COUNCILSSurvey respondents selected what they thought should bethe “main role” of City Boards and Neighborhood Councils. BOARD/COUNCIL INFLUENCESurvey respondents selected how much influence theythought these entities should have on city decisions. 0 50 100 150 200 250 ape policies, limited decision-making power ns to Commission, not make final decisions ng power, approve or block certain actions Unsure Share input and shape policies,limited decision-making power Provide advice to Commission, notmake final decisions (current form) Formal decision-making power,approve or block certain actions Unsure 0 50 100 150 200 250 Shared influence Advisory only Strong influence Unsure Shared Influence (Boards help setpriorities and may have authority incertain areas) Advisory Only (Commission mayconsider input, but makes all decisions-current form) Strong Influence (Boards can veto orblock decisions made by Commission) Unsure DEMOGRAPHIC PROFILE ZIP CODERespondents were primarily from 59715.QUADRANTThe SE and NW quadrants were moststrongly represented in this data. RESIDENCY STATUSMost respondents were full-time residents(more than 6 months per year). 0 50 100 150 200 250 300 350 59715 59718 59717 0 50 100 150 200 SE NW NE SW Outside City Limits 0 100 200 300 400 500 600 Full-time resident Other Part-time resident LOCATIONMost respondents lived within city limits.0100200300400500Within Bozeman city limits Outside city limits in Gallatin Valley STUDENT STATUSMost respondents were not students. 0 100 200 300 400 500 Not a Student Student at MSU Student at Gallatin College AGE CATEGORIES Survey respondents (average = 55 yrs)skewed older than census populationestimates (average = 35 yrs). Census Survey Working Age (16-64)Older Individuals (65+)0% 20% 40% 60% 80% 100% GENDERSurvey respondents included a slightoverrepresentation of women (54%)compared to census populationestimates (47%). Census Survey Female/WomanMale/ManNon-binaryPrefer not to respondPrefer to self-describe0% 10% 20% 30% 40% 50% 60% RACE/ETHNICITYSurvey respondents mirrored theracial/ethnic profile in censuspopulation estimates. Census Survey WhiteHispanic/LatinoAsianAmerican Indian/Alaskan NativeBlack/African AmericanNative Hawaiian/Pacific Islander0% 20% 40% 60% 80% 100% Note. Additional exploratory analyses by age did not meaningfully change results or conclusions. Community Survey 18 6 Strongly Disagree Strongly Agree Community Employee 1 2 3 4 5 I stay informed about local government issues in Bozeman. I understand who leads and makes decisions for our city. I am confident in the current structure of city government in Bozeman. Structure enables clear roles, responsibilities, and reporting lines for city employees. Structure supports efficient delivery of services in my department CURRENT STATE PERCEPTIONS RESPONDENTS FEEL INFORMED ABOUT THE CITY, BUT HAVE LOWER CONFIDENCE AND MIXED PERCPETIONS OF STRUCTURE Understanding & AttitudesSurvey respondents rated items on a scale from 1 (strongly disagree) to 5(strongly agree). LEADERSHIP PRIORITIES TOP PRIORITIES FOR EMPLOYEES INCLUDE ACCOUNTABILITY, TRANSPARENCY, EFFICIENT ADMINISTRATION, AND STABLEADMINISTRATION 0 50 100 150 200 250 300 Accountable Leadership Transparent Leadership Responsive Administration Fiscally-Responsible Model of Government Democratic Representation Efficient Administration Stable Administration Term-Limited Leaders Visible Leaders Other COMMUNITY SURVEYSurvey respondents voted on their top 3 prioritiesfor how they want the City of Bozeman leadership tofunction going forward. EMPLOYEE SURVEY Survey respondents voted on their top 3 priorities forhow they want the City of Bozeman leadership tofunction going forward. 0 20 40 60 80 100 Accountable Leadership Transparent Leadership Responsive Administration Fiscally-Responsible Model of Government Democratic Representation Efficient Administration Stable Administration Term-Limited Leaders Visible Leaders Other Employee Survey 19 7 IMPACT THE QUESTION: On a scale of 1-5, what level of impact do each of the following roles have on your day-to-day work? When responding,please focus on the roles, responsibilities, and structure of city leadership positions rather than the individuals currentlyserving in those roles. No Impact Very Significant Impact Average Employee Rating 1 2 3 4 5 Mayor Deputy Mayor City Commissioners City Manager Neighborhood Councils City Boards 2.2 1.97 2.7 3.87 1.91 2.17 MAYOR FEEDBACK COMMUNITY SURVEYSurvey respondents selected howinvolved the mayor should be in dailycity operations. Hybrid Executive 33.8% Legislative Leader 31.3% Executive Leader18.5% Hybrid Legislative 13.7% EMPLOYEE SURVEYSurvey respondents selected howinvolved the mayor should be in dailycity operations. Legislative Leader 60% Hybrid Legislative 21.2% Hybrid Executive 12.1% Executive Leader4.2% SURVEY OPTIONS Hybrid Executive: Mayor handles some executive tasks, but City Manager still manages most daily operations Legislative Leader: Mayor leads commission meetings and represents city, but does not run departments (current form) Executive Leader: Mayor elected separately, directly manages city operations Hybrid Legislative: Mayor breaks ties in commission meetings, but does not manage day-to-day government Employee Survey 20 8 EMPLOYEE SURVEYSurvey respondents selected whetherthe city should retain a professionalcity manager. CITY MANAGER FEEDBACK Yes, FullExecutiveAuthority 72% Yes, Reduced powers 18.9% No, MayorExecutiveAuthority 5.5%Yes, Reduced Powers: City manager handles some operations, while the Mayor takes on more authority Yes, Full Executive Authority: City Manager runs all operations; Mayor is mostly legislatively focused (current form) No, Mayor Executive Authority: The structure should allow the Mayor to hold all the executive authority Unsure Yes, Reduced powers 50.5% Yes, FullExecutive Authority 34.6% No, MayorExecutiveAuthority 11.7% Unsure3.3% COMMUNITY SURVEY Survey respondents selected whetherthe city should retain a professionalcity manager.SURVEY OPTIONS ELECT ADDITIONAL OFFICIALS 0 50 100 150 200 250 Yes, I would support electing more city officials No, I would prefer these positions remain appointed Not sure / Need more information COMMUNITY SURVEY How the Question WasFramed: Survey respondents selectedwhether they would support theidea of having additional cityofficials elected by voters. Forexample, in some cities, positionssuch as City Clerk, City Treasurer,or City Attorney are elected. EMPLOYEE SURVEY 0 20 40 60 80 100 120 Yes, I would support electing more city officials No, I would prefer these positions remain appointed Not sure / Need more information Employee Survey 21 0 20 40 60 80 100 Commissioners live in district, Voters only from district Commissioners live in district, Voters from across the city Commissioners live across city, Voters only from district No preference -- 10 20 30 40 50 1- Strongly Disagree 2- Disagree 3- Neutral 4- Agree 5- Strongly Agree 23 20 45 45 31 No53.4% Unsure23.9% Yes22.7% 9 GEOGRAPHIC VS. CITY-WIDE COMMISSIONERSSurvey respondents rated their preference on a scalefrom 1 (strongly disagree) to 5 (strongly agree). CITY COMMISSION FEEDBACK POTENTIAL GEOGRAPHIC RULESEmployee respondents also considered IF geographic-basedcommissioners were used, which rules would make the most sense. I believe Bozeman should prioritize geographic district- based representation rather than total city-wide votes for commissioners. COMMISSION EXPANSIONSurvey respondents also selected whether they thought Bozemanshould expand its City Commission beyond 5 members. Employees Survey No41.7% Yes33.7% Unsure24.6% Community Survey ADVISORY BOARDS & NEIGHBORHOOD COUNCILS FEEDBACK ADVISORY BOARDS & NEIGHBORHOOD COUNCILSSurvey respondents selected what they thought should bethe “main role” of City Boards and Neighborhood Councils. 0 20 40 60 80 100 d shape policies, limited decision-making power ations to Commission, not make final decisions making power, approve or block certain actions Unsure Share input and shape policies, limited decision-making power Provide advice to Commission, not make final decisions (current form) Formal decision-making power, approve or block certain actions Unsure BOARD/COUNCIL INFLUENCESurvey respondents selected how much influence theythought these entities should have on city decisions. 0 20 40 60 80 100 Shared influence Advisory only Strong influence Unsure Shared Influence (Boards help setpriorities and may have authority incertain areas) Advisory Only (Commission mayconsider input, but makes all decisions-current form) Strong Influence (Boards can veto orblock decisions made by Commission) Unsure Employee Survey 22 10 CITY RESIDENCYMost respondents lived within city limits. DEMOGRAPHIC PROFILE RESPONDENTS WERE PRIMARILY CITY RESIDENTS WORK CATEGORYThe category of infrastructure and Utilities wasthe top choice. LENGTH OF EMPLOYMENT A variety of responses in this category020406080100Within Bozeman city limits Outside city limits 0 10 20 30 40 50 Infrastructure and Utilities Internal Administration Public Safety Community and Resident Services Planning and Development Other 0 20 40 60 80 Less than 5 years 5-10 years More than 10 years 0 20 40 60 80 100 Yes No SUPERVISORY/MANAGEMENT ROLEMost respondents were not in managementroles. Employee Survey 23 11 Strongly Disagree Strongly Agree Community Employee Board 1 2 3 4 5 I stay informed about local government issues in Bozeman. I understand who leads and makes decisions for our city. I am confident in the current structure of city government in Bozeman. Structure enables clear roles, responsibilities, and reporting lines for city employees. Structure supports efficient delivery of services in my department CURRENT STATE PERCEPTIONS Understanding & AttitudesSurvey respondents rated items on a scale from 1 (strongly disagree) to 5(strongly agree). LEADERSHIP PRIORITIES Survey respondents voted on their top 3 priorities forhow they want the City of Bozeman leadership tofunction going forward. BOARD SURVEY 0 5 10 15 20 25 Accountable Leadership Transparent Leadership Responsive Administration Fiscally-Responsible Model of Government Democratic Representation Efficient Administration Stable Administration Term-Limited Leaders Visible Leaders Other COMMUNITY SURVEY 0 50 100 150 200 250 300 Accountable Leadership Transparent Leadership Responsive Administration Fiscally-Responsible Model of Government Democratic Representation Efficient Administration Stable Administration Term-Limited Leaders Visible Leaders Other EMPLOYEE SURVEY 0 20 40 60 80 100 Accountable Leadership Transparent Leadership Responsive Administration Fiscally-Responsible Model of Government Democratic Representation Efficient Administration Stable Administration Term-Limited Leaders Visible Leaders Other City Board Survey 24 12 ROLE AS BOARD MEMBER WE ASKED ABOUT THEIR ROLES AS BOARD MEMBERS APPROXIMATELY HOW MANY HOURS EACH WEEK DOES ITTAKE YOU TO DO YOUR WORK FOR THE BOARD YOU SERVE ON? DO YOU FEEL YOUR IDEAS AREBEING LISTENED TO WHILESERVING ON THE BOARD? ARE YOU BEING PROPERLYCONSULTED FOR YOURKNOWLEDGE AND EXPERTISEWHILE SERVING ON THE BOARD? DO YOU FEEL CONSTRAINED INOPENLY SHARING IDEAS WHILESERVING ON THE BOARD 0 10 20 2 or less 2.1-4.0 4.1-6.0 Above 6 0 5 10 15 20 25 30 Yes No 0 5 10 15 20 25 30 Yes No 0 5 10 15 20 25 30 35 Yes No ROLE AS BOARD MEMBER WE ASKED ABOUT THEIR ROLES AS BOARD MEMBERS 0 10 20 30 too broad too narrow just right IS THE SCOPE OF THE BOARD YOU SERVE ON... IS THE BOARD YOU SERVE ONSPECIFIC ENOUGH TO ITS SCOPETO ADDEQUATLEY UTILIZE YOURKNOWLEDGE AND INTERESTS? 0 10 20 30 40 Yes No SHOULD CITY BOARDS HAVETERM LIMITS? 0 5 10 15 20 25 30 Yes No City Board Survey 25 -- 2 4 6 8 10 12 14 1- Strongly Disagree 2- Disagree 3- Neutral 4- Agree 5- Strongly Agree 11 6 4 12 10 13 GEOGRAPHIC VS. CITY-WIDE COMMISSIONERSSurvey respondents rated their preference on a scalefrom 1 (strongly disagree) to 5 (strongly agree). GEOGRAPHY + CITY COMMISSION The Question: I believe Bozeman should prioritize geographic district-based representation rather than total city-wide votes for commissioners. EMPLOYEE SURVEY COMMUNITY SURVEY BOARD SURVEY -- 10 20 30 40 50 1- Strongly Disagree 2- Disagree 3- Neutral 4- Agree 5- Strongly Agree 23 20 45 45 31 -- 50 100 150 200 1- Strongly Disagree 2- Disagree 3- Neutral 4- Agree 5- Strongly Agree 88 63 84 160 141 CITY COMMISSION FEEDBACK COMMISSION EXPANSIONSurvey respondents also selected whether they thought Bozemanshould expand its City Commission beyond 5 members. *unsure option removed for Board Survey No 57.8% Yes42.2% BOARD SURVEY No41.7% Yes 33.7% Unsure24.6% COMMUNITY SURVEY No53.4% Unsure23.9% Yes22.7% EMPLOYEE SURVEY City Board Survey 26 SURVEY OPTIONS Survey respondents selected howinvolved the mayor should be in dailycity operations. 14 MAYOR FEEDBACK Hybrid Executive 33.8% Legislative Leader 31.3% Executive Leader18.5% Hybrid Legislative 13.7% Hybrid Executive: Mayor handles some executive tasks, but City Manager still manages most daily operations Legislative Leader: Mayor leads commission meetings and represents city, but does not run departments (current form) Executive Leader: Mayor elected separately, directly manages city operations Hybrid Legislative: Mayor breaks ties in commission meetings, but does not manage day-to-day government Legislative Leader 60% Hybrid Legislative 21.2% Hybrid Executive 12.1% Executive Leader4.2% EMPLOYEE SURVEY COMMUNITY SURVEYBOARD SURVEY Legislative Leader 36% Hybrid Executive 30% Executive Leader16% Hybrid Legislative 10% CITY MANAGER FEEDBACK SURVEY OPTIONS Survey respondents selected whetherthe city should retain a professionalcity manager. Yes, Reduced Powers: City manager handles some operations, while the Mayor takes on more authority Yes, Full Executive Authority: City Manager runs all operations; Mayor is mostly legislatively focused (current form) No, Mayor Executive Authority: The structure should allow the Mayor to hold all the executive authority Unsure Yes, Full Executive Authority 54.3% Yes, Reduced powers 32.6% No, Mayor ExecutiveAuthority 10.9% Unsure 2.2% BOARD SURVEY Yes, Reduced powers 50.5% Yes, FullExecutive Authority 34.6% No, MayorExecutiveAuthority 11.7% Unsure3.3% COMMUNITY SURVEY Yes, Full Executive Authority72% Yes, Reduced powers 18.9% No, Mayor Executive Authority 5.5% EMPLOYEE SURVEY. City Board Survey 27 ELECT ADDITIONAL OFFICIALS COMMUNITY SURVEY 0 50 100 150 200 250 Yes, I would support electing more city officials No, I would prefer these positions remain appointed Not sure / Need more information EMPLOYEE SURVEY 0 20 40 60 80 100 120 Yes, I would support electing more city officials No, I would prefer these positions remain appointed Not sure / Need more information How the Question Was Framed: Survey respondents selected whether they wouldsupport the idea of having additional city officialselected by voters. For example, in some cities,positions such as City Clerk, City Treasurer, or CityAttorney are elected. *Not Sure Option not included on Board Survey BOARD SURVEY 0 5 10 15 20 25 30 35 Yes, I would support electing more city officials No, I would prefer these positions remain appointed 15 BOARD SURVEY RESULTS ADVISORY BOARDS FEEDBACK 0 5 10 15 20 25 shape policies, limited decision-making power ations to Commission, not make final decisions making power, approve or block certain actions Share input and shape policies,limited decision-making power Provide advice to Commission, not make final decisions (current form) Formal decision-making power,approve or block certain actions 0 5 10 15 20 25 Shared influence Advisory only Strong influence Shared Influence (Boards help setpriorities and may have authority in certain areas) Advisory Only (Commission mayconsider input, but makes all decisions-current form) Strong Influence (Boards can veto or block decisions made by Commission) ADVISORY BOARD INFLUENCESurvey respondents selected how much influence theythought these entities should have on city decisions. ADVISORY BOARDS ROLESurvey respondents selected what they thought should bethe “main role” of City Boards. City Board Survey 28 16 NEIGHBORHOOD COUNCILS FEEDBACK NEIGHBORHOOD COUNCILSSurvey respondents selected what they thought should bethe “main role” of Neighborhood Councils. NEIGHBORHOOD COUNCIL INFLUENCESurvey respondents selected how much influence theythought these entities should have on city decisions. 0 5 10 15 20 25 e policies, limited decision-making power to Commission, not make final decisions power, approve or block certain actions Share input and shape policies,limited decision-making power Provide advice to Commission, notmake final decisions (current form) Formal decision-making power, approve or block certain actions 0 5 10 15 20 25 30 Shared influence Advisory only Strong influence Shared Influence (Boards help set priorities and may have authority incertain areas) Advisory Only (Commission mayconsider input, but makes all decisions-current form) Strong Influence (Boards can veto orblock decisions made by Commission) DEMOGRAPHIC PROFILE Experience and Role 01020304050Yes No HAVE YOU PREVIOUSLY HELD ANELECTED OFFICE? HAVE YOU PREVIOUSLY OR DO YOUCURRENTLY WORK FOR THE CITY OFBOZEMAN? 0 10 20 30 40 50 Yes No HOW LONG HAVE YOU LIVED INBOZEMAN? 0 10 20 30 40 Less than 5 years 5-10 years More than 10 years HOW LONG HAVE YOUR SERVED ONA BOARD? 0 5 10 15 20 25 Less than 5 years 5-10 years More than 10 years City Board Survey 29 Memorandum REPORT TO:Study Commission FROM:Caeleb Heinen, Recording Secretary Mike Maas, Ex Officio Dan Clark, MSU Local Government Center SUBJECT:Further Learning for Wards/Districts MEETING DATE: April 2, 2026 AGENDA ITEM TYPE: Citizen Advisory Board/Commission RECOMMENDATION: The Study Commissioners and Ex Officio requested more data and materials on certain topics from their previous Learning Session on Wards and Districts. Dan Clark from MSU Local Government Center has provided more research to cover their interests. For the March 25th Reading Materials, click here. STRATEGIC PLAN: 1.1 Outreach: Continue to strengthen and innovate in how we deliver information to the community and our partners. BACKGROUND: NEW Reading on Voter Turnout and Gerrymandering: Comparison City Chart (with Voter Turnout) Voter Turnout in At Large vs. Wards Correlates of Voter Turnout - Political Behavior Seattle Districting Commission San Diego Charter Redistricting March 25th Agenda Reading Materials Mandatory Reading: At-Large VS Wards Wards Documents for Bozeman 2026 Supplemental Information (See March 25th Agenda): District Based Elections and Class Based Representation Evidence from the California Voting Rights Act Free Cities Center Analysis City Representation MIT At Large Elections and Minority Representation in Local Government Primer Districts vs. At-Large Elections Digital Trade-Offs between at-large and single member districts UofHouston Wards at-large in Canadian Cities UNRESOLVED ISSUES: 30 Wards vs At Large on Diversity At-Large vs Ward - Implications for Public Infrastructure Hybrid City Council Election Systems None Identified ALTERNATIVES: As per the Study Commission FISCAL EFFECTS: None Identified Attachments: Comparison City Chart (with Voter Turnout).pdf Voter Turnout in At Large vs. Wards.pdf Correlates of Voter Turnout Political Behavior.pdf Seattle Districting Commission.pdf San Diego Charter Redistricting.pdf Report compiled on: March 25, 2026 31 Government Characteristics Anaconda- Deer Lodge Butte-Silver Bow Belgrade Billings Bozeman Columbia Falls Great Falls Havre Helena Kalispell Lewistown Livingston Missoula Polson West Yellowstone Whitefish Santa Fe Seattle Eugene At-Large/Wards (AL/W)5 Districts 12 Districts 3 Wards 5 Wards At-Large At-Large At-Large 4 Wards At-Large 4 Wards 3 Wards + 1 AL At-Large 6 Wards 3 Wards At-Large At-Large 4 Wards 9 Wards 2 AL 4 Council Terms 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years 4 Years Number of Council Members + Mayor/PO as part of Legislative Branch 5 12 6+M 10+M 4+M 6+M 4+M 8 4+M 9+M 7/PO*5/PO*12 6+M 5/M*6+M 8+M 9 8 Mayor/Presiding Officer (CEO/Mayor/PO)CEO CEO Mayor Mayor Mayor Mayor Mayor Mayor Mayor Mayor PO PO Mayor Mayor Mayor Mayor Mayor Mayor Mayor Mayor/Deputy Mayor Split Term (Y/N)No No No No Yes No No No No No No No No No No No No No No Mayor/P.O. Term of Office 4 yr 4 yr 2 yr 4 yr 2+2 4 yr 2 yr 4 yr 4 yr 4 yr 4 yr 4 yr 4 yr 4 yr 1 yr 4 yr 4 yr 4 yr 4 yr Com-Mgr/Com-Ex Structure Com-Ex Com-Ex Com-Mgr Com-Mgr Com-Mgr Com-Mgr Com-Mgr Com-Ex Com-Mgr Com-Mgr Com-Mgr Com-Mgr Com-Ex Com-Mgr Com-Mgr Com-Mgr Com-Ex w CAO Com-Ex Com-Mgr SG Charter/General Gov Charter Charter Charter Charter Charter Gen Gov Charter Charter Charter Gen Gov Charter Gen Gov Charter Charter Charter Charter Charter Charter Charter EO Vacancy**120 days 60 days 60 Days 30 days 30 days 30 days 30 days**30 days**30 days**30 days 30 days 30 days 30 days 30 days 30 days 30 days 30 days 20 90 Neighborhood Councils No Kind of No No Yes No Yes No Yes No No No Yes No No No Voter Turnout 2025 34%47%34%54%52%NA 42% Voter Turnout 2023 26%29%26%NA 23%NA 45% *Mayor is elected by the council from its membership **Charter states "as may be provided by law", does not specify a timeline (and MCA defaults) or the commission has passed an ordiance outlining a specific timeline for filling an EO position. 32 Voter Turnout in At-Large vs. Ward-Based Elections To meet the Study Commission’s request for a rapid assessment of how at-large versus ward-based election systems affect voter turnout, I used artificial intelligence to synthesize peer-reviewed academic articles previously referenced in discussions of election structure, along with additional relevant research. The following summarizes the results of that synthesis. Overall, the research finds that ward-based election systems do not consistently increase total voter turnout. However, they do reduce turnout inequality and increase meaningful participation among communities that typically vote at lower rates. Theoretical Expectations Political participation theory indicates that district magnitude and perceived voter influence play an important role in shaping voter turnout. Ward systems reduce constituency size, which tends to make it: • Easier for voters to believe their vote matters, • Easier for candidates to engage in face-to-face and neighborhood-level campaigning, • Easier for voters to identify “their” representative. Classic rational-choice and voter mobilization theories predict higher participation where information costs are lower and mobilization is more targeted (Downs, 1957; Rosenstone & Hansen, 1993). These dynamics align more naturally with ward-based elections than with citywide at-large elections. Although these expectations originate in theory, they have been repeatedly examined and supported in empirical research on local government elections. Evidence from U.S. Municipal Elections The strongest evidence from the United States comes from comparative studies that exploit variation in electoral institutions across cities. Ferraz and Finan (2011) find that elections with smaller constituencies and clearer lines of accountability produce higher participation and political engagement. Subsequent research extends these findings to U.S. municipal contexts, where the effects are strongest in low-information elections such as city council contests. Trounstine (2018) and related studies show that at-large systems are associated with lower participation among geographically concentrated groups, in part because turnout incentives are diluted within large, citywide electorates. While this work focuses primarily on representation, turnout inequality consistently appears as an underlying mechanism. 33 Historical analyses further show that many at-large systems—particularly in the U.S. South—were adopted during periods explicitly aimed at restricting participation by certain groups, contributing to long-term disparities in political engagement. Conclusions for the United States Taken together, peer-reviewed research generally finds equal or lower turnout under at-large systems, with ward-based systems producing higher participation among voters whose turnout is otherwise fragile. The strongest consensus in the literature is that ward-based municipal elections tend to produce modestly higher voter turnout among lower-income and minority voters, though this effect is conditional rather than universal. The turnout advantages of ward systems are most evident when: • Elections are low-salience and receive limited public attention and engagement compared with high-salience elections such as presidential or gubernatorial races and are often stand-alone; • Campaigns are less expensive and more localized; • Voters perceive clearer geographic accountability and responsiveness. At the same time, the research consistently emphasizes that election structure alone does not determine turnout. Election timing, competitiveness, and broader social and political context often matter as much as and often or more than the choice between at-large and ward-based systems. 34 Vol.:(0123456789) Political Behavior (2023) 45:607–633 https://doi.org/10.1007/s11109-021-09720-y 1 3 ORIGINAL PAPER Correlates of Voter Turnout Richard W. Frank1  · Ferran Martínez i Coma2 Accepted: 11 May 2021 / Published online: 20 May 2021 © The Author(s) 2021 Abstract Despite decades of research, there is no consensus as to the core correlates of national-level voter turnout. We argue that this is, in part, due to the lack of com- prehensive, systematic empirical analysis. This paper conducts such an analysis. We identify 44 articles on turnout from 1986 to 2017. These articles include over 127 potential predictors of voter turnout, and we collect data on seventy of these vari- ables. Using extreme bounds analysis, we run over 15 million regressions to deter- mine which of these 70 variables are robustly associated with voter turnout in 579 elections in 80 democracies from 1945 to 2014. Overall, 22 variables are robustly associated with voter turnout, including compulsory voting, concurrent elections, competitive elections, inflation, previous turnout, and economic globalization. Keywords Elections · Turnout · Extreme bounds analysis · Meta-analysis Introduction A common challenge in the study of comparative politics is balancing theoretical and empirical comprehensiveness with substantive importance. Consider voter turn- out. If we ask what the most statistically significant and substantively important pre- dictors of national-level voter turnout in democratic elections are, even after more than 50 years of comparative voter turnout research, there are few certainties beyond the fact that compulsory voting increases turnout. For example, several studies including Radcliff and Davis (2000) find larger district magnitudes increase turnout while others like Tavits (2008) find either no significant relationship or even a nega- tive one (Fumagalli & Narciso, 2012). * Richard W. Frank richard.frank@anu.edu.au 1 School of Politics and International Relations, Australian National University, Canberra, Australia 2 People, Elections and Parties Research Group, Centre for Governance and Public Policy, School of Government and International Relations, Griffith University, Brisbane, Australia 35 608 Political Behavior (2023) 45:607–633 1 3 One possible reason for these sorts of contradictory findings is that a topic has not received enough research attention for a consensus to emerge. This is not the case for voter turnout; it is one of the most studied topics in the discipline. Indeed, Cancela and Geys (2016, p. 264) suggest “turnout scholarship witnessed a verita- ble explosion” in the last 15 years. A second possibility is that we lack a thorough understanding of the underlying explanatory factors. Again, this does not seem to be the case here given the profusion of turnout meta-analyses. A recent meta-analy- sis of 130 journal articles identifies over a hundred explanatory factors (Stockemer, 2017). Geys’s (2006) earlier meta-analysis of 83 studies focuses on fourteen corre- lates but identifies several dozen more; and more recently Cancela and Geys (2016) examine 102 studies and identify several dozen correlates. Therefore, although there are myriad possible factors driving voter turnout, it has been difficult to reach solid empirical conclusions. A third possibility is that the world’s contextual heterogene- ity explains why some variables behave differently in certain contexts, driving con- tradictory findings. Although empirical conflicts can definitely arise from contextual differences, they do not tell the whole story. For if a goal of comparative politics is reaching solid and generalizable conclusions across contexts, it is important to sys- tematically approach competing explanations for comparable outcomes while recog- nizing important contextual differences. The comparable outcome we explore here is national-level voter turnout. In the national-level voter turnout literature, it is uncommon to claim that one empirical model trumps others (Temple, 2000). However, there are developed tech- niques to systematically evaluate the proposed factors for a political outcome includ- ing meta-analyses and extreme bounds analyses. Given the sizable, established literature on voter turnout, this paper’s contribution is synthesizing the recent litera- ture and evaluating its proposed correlates of national-level voter turnout using an extreme bounds analysis. Extreme bounds analysis (EBA) has been used in a wide variety of contexts to evaluate factors driving a number of political and economic outcomes (Leamer, 1983; Levine & Renelt, 1992; Sala-i-Martin, 1997). For example, it has been used to evaluate over 50 predictors of economic growth (Levine & Renelt, 1992), 20 pos- sible factors contributing to human rights violations (Hafner-Burton, 2005), 59 pre- dictors of democracy (Gassebner, Lamla, & Vreeland, 2013), 59 electoral integrity predictors (Frank & Martínez i Coma, 2017), 43 covariates of life expectancy and infant mortality (Carmignani et al., 2014), 53 determinants of health care expendi- tures (Hartwig & Sturm, 2014), and 23 factors behind the diffusion of coups (Miller, Joseph, & Ohl, 2018). While many proposed proxies in these areas are statistically significant when considered in isolation, when tested with other predictors such findings are often fragile (Leamer, 1983). Extreme bounds analysis allows us to sys- tematically evaluate what factors are robust to different model specifications. Hence, a primary EBA goal is to show that the “assumed model specification is largely inconsequential for statistical inference” (Gassebner, Gutmann, & Voigt, 2016, p. 295). Another goal is to reconcile the literature’s several (sometimes contradictory) findings. To collect possible predictors of national-level electoral turnout, we analyze 44 articles on voter turnout published between 1986 and 2017 in leading political 36 609 1 3 Political Behavior (2023) 45:607–633 science journals. We identified 127 unique independent variables that may affect turnout, and we were able to collect data and run models using seventy of these vari- ables in 579 elections in 80 democracies from 1945 to 2014. We then run over two million regressions with different combinations of these seventy predictors. Each variable was included in up to 41,660 models with various combinations of other variables. If a variable is a significant predictor across models, then we can conclude that its statistical significance is unlikely to be an artefact of model specification. To determine robust turnout predictors, we used the two most common EBA decision criteria proposed by Leamer (1983) and Sala-i-Martin (1997), and we find 7 and 22 variables respectively are robust according to these 2 sets of criteria across 2 model specifications.1 The first series of models includes country fixed effects to control for unobserved country-level factors; the second series of models includes random effects which allows for the inclusion of sluggish or stationary country-level factors the literature suggests affects turnout. We also run a number of further sensitivity analyses excluding a lagged dependent variable, using a dependent variable with a different denominator (voting age population rather than registered voters), and run- ning models on eight election subsamples. This research is theoretically significant because we still lack a systematic and parsimonious explanation of voter turnout that can address the current inconclusive and sometimes contradictory nature of the literature’s empirical results. The aim of this paper is, therefore, threefold: (1) to shed light on the dozens of factors that can affect turnout; (2) assess the empirical robustness of the different explanatory fac- tors; and (3) provide insight on which controls may be worth including in future work on voter turnout.2 We proceed as follows. The next section briefly summa- rizes the current voter turnout literature. The third section examines meta-analysis’s strengths and weaknesses and describes how extreme bounds analysis compliments it. The research design section discusses our election sample, the dependent and independent variables, and several estimation considerations. Our main results are then presented and are followed by a series of further analyses. We then conclude with a discussion of our main findings and areas for future research. What Do We Know About Voter Turnout? The first national-level turnout studies explain variations in voter turnout by focus- ing on a selected sample of lower house elections in OECD countries. For example, Jackman (1987) analyzes 19 democracies, Jackman and Miller (1995) analyze 23, and Blais and Carty (1990) and Powell (1986) include 20. Normally, these studies 1 Using extreme bounds analysis, we examine the most robust factors correlated with turnout. Like pre-vious studies using this methodology, we do not estimate a structural model, theorize the relationship between different variables, establish specific causal mechanisms, or improve turnout measurements.2 Existing meta-analyses present contradictory results. In fact, Stockemer (2017, p. 15) states “the fact that the influence of many factors … on turnout is inconclusive demands more contextual analysis.” While we agree with the need for context bound analysis, our research shows that even at the most gen-eral level there are common voter turnout correlates in democratic countries. 37 610 Political Behavior (2023) 45:607–633 1 3 focus on a series of factors such as “socio-economic environment, the constitutional setting, and the party system,” (Blais & Dobrzynska, 1998, p. 241). Recent years have seen a proliferation of potential theoretical factors shaping turnout as well as an expansion of coverage. For instance, voter turnout is depressed with the size of a political community (Remmer, 2010), economic globalization (Steiner, 2010), corruption (Stockemer, LaMontagne, & Scruggs, 2013), and ethnic diversity (Martínez i Coma & Nai, 2017). Terrorism, in turn, increases the elector- ates’ attention on national politics and, consequently, turnout increases (Robbins, Hunter, & Murray, 2013). Similarly, while previous studies focused mostly on eco- nomically developed democracies, Blais and Dobrzynska (1998) were the first to have a truly global collection of democratic elections. More recent analyses focus on other regions including Latin America (Fornos, Timothy Power, & Garand, 2004), post-Cold War Eastern Europe (Kostadinova, 2003), Africa (Kuenzi & Lambright, 2007), and Muslim-majority countries (Stockemer & Khazaei, 2014). These works share a number of similarities including the use of three general types of independent variables: socioeconomic, institutional, and political (Geys, 2006; Blais & Dobrzyinska, 1998). Socioeconomic variables include factors like economic growth and urbanization. Institutional variables capture the institutional arrangements under which the election take place (e.g., electoral formula). Finally, political variables account for specific electoral outcomes (e.g., the margin of victory).3 The literature’s limitations appear when comparing their results. For example, one of the most cited studies by Blais and Dobrzyinska (1998) find that turnout is significantly affected by economic development, literacy rates, population size and density, compulsory voting laws, minimum voting ages, the electoral system, the number of political parties seeking seats, and the election’s competitiveness. End- ersby and Krieckhaus (2008) reach similar empirical conclusions but suggest that context is important. Along the same lines, Martínez i Coma (2016) confirms some of Blais and Dobrzyinska’s (1998) findings but not others—notably those relating to electoral systems and economic development. Furthermore, and more important for the purpose of this paper, a consensus does not yet exist on the robustness of these variables and, consequently, on what vari- ables should be considered for a core model of cross-national aggregate voter turn- out. For example, of the eight socio-economic factors used in the three articles men- tioned above, only one variable is considered in all three—population.4 Put simply, comparative studies provide mixed evidence for the robustness of particular factors affecting voter turnout. In summary, after over 50 years of research the literature still has not coalesced around a core model of turnout; different sets of variables are used in different analysis.5 The literature, “draw[s] on relatively small samples, differing 3 Such categorizations are neither exhaustive nor exclusive; rather they can be seen as a useful theoreti- cal heuristic.4 The other variables are population density, gross domestic (or national) product per capita, gross domestic product, literacy rate, life expectancy, and ethnic diversity.5 Even when studies include the same variables, there is no consensus on how to measure some con-cepts. See below for a broader measurement discussion. 38 611 1 3 Political Behavior (2023) 45:607–633 sets of observations, divergent operationalizations of turnout, and a wide array of theoretical models, the literature has mainly converged around relatively common- sensical observations” (Remmer, 2010, p. 277). Why a Meta‑analysis is Not Enough One popular means of evaluating a literature’s lessons is a meta-analysis, which basically assesses whether (or not) an independent variable affects a dependent vari- able.6 A meta-analysis, or an “analysis of analyses” (Glass, 1976, p. 3), normally employs two procedures: “vote-counting” and “combined tests.” The former counts the number of times a given coefficient is significant and in the expected direction. In such cases, it is counted as a ‘success’; otherwise, it is considered an anomaly. The higher the success rate, the more likely it is that we are to be observing a real association between the variables. A combined test is “based on the summation of the actual test statistics provided in each study” (Geys, 2006, p. 640).7 Such meta-analyses have been done in the voter turnout literature. For example, Geys (2006, p. 641) identifies twelve significant socio-economic, political, and insti- tutional factors affecting turnout, while highlighting that none are “omnipresent in the literature.”8 This is “partly due to the lack of a firm theoretical model at the basis of variable selection,” and he argues for the “construction of a ‘core’ model of turnout” (Geys, 2006, pp. 641, 653). Along similar lines, Stockemer’s (2017, p. 712) meta-analysis of 135 studies from 2004 to 2013 identifies over 100 potentially salient variables, thoroughly analyzes 10 of them, and concludes “no variable is omnipresent or appears in most studies. Rather, different variables are used in vari- ous contexts.” The divergence pointed out by Stockemer (2017) explains the differ- ent results from his and Geys’ (2006) meta-analysis. They assess the “success” or robustness of 18 variables, 5 of those common in both works. Even for those five common variables,9 when comparing the studies’ success rates, there are three com- mon results: compulsory voting and population size impact turnout while income inequality does not. They diverge on the impact of election closeness and PR system because Geys (2006) finds they affect turnout, while Stockemer (2017) does not. Unsurprisingly then, Stockemer (2017, p. 712) acknowledges that the “literature is far from establishing a core turnout model.” 6 In discussing meta-analyses, we are not referring to quantitative analysis of a variable’s average treat- ment effect because this literature’s focus is on a broad spectrum of possible causes rather than any one particular cause. 7 A limitation of such an approach is it requires consistent reporting across the studies, but there are solutions. For example, see Geys (2006) and Smets and van Ham (2013). 8 Cancela and Geys (2016) expand on Geys’ (2006) by adding 102 new studies and differentiating between national and subnational elections. 9 The variables that only Geys (2006) considers are: population concentration, population stability, population homogeneity (ethnic diversity), previous turnout, campaign expenditures, political fragmen-tation, proportional representation electoral system, concurrent elections, and registration requirements. The variables that only Stockhemer (2017) considers are district magnitude, effective number of parties, important elections, education, and literacy rate. 39 612 Political Behavior (2023) 45:607–633 1 3 Meta-analysis’s main limitation is not procedural but conceptual. In the end, even the most exhaustive meta-analyses like those above are circumscribed by their sam- ples. This implies that we are unsure about the robustness of such results, given that most researchers’ robustness checks are ad hoc. “They identify a set of competing explanations and see if their empirical results hold once they control for some vari- ables that might be consistent with those explanations” (Hegre & Sambanis, 2006, p. 509). In other words, given all possible variable combinations, we cannot be sure that the selected model and the results presented are the ‘right’ ones. Usually, sensitivity (or robustness) checks estimate a series of regressions with alternative specifications. However, the number of possible alternative specifications is, of course, substantially higher. For example, without assuming any initial knowledge of variable selection, 5 variables lead to 32 possible model specifications, 6 variables have 64 combinations, 7 variables have 5040, 8 have 40,320, etcetera. The number of permutations increases exponentially as the number of variables increases; therefore, any particular study is likely showing a tiny proportion of the possible combinations. Hence, even the most demanding and detailed meta-analysis will only cover a tiny percentage of the multiple possible combinations that may affect voter turnout. The standard regression framework on which meta-analyses rely has two other limi- tations. First, a particular variable’s statistical significance may be sensitive to the inclu- sion/exclusion of other variables. As Leamer (1983, p. 38) concludes, “an inference is not believable if it is fragile, if it can be reversed by minor changes in assumptions.” Second, even when theories point to particular mechanisms, they are not “refined enough to inform the choice of the empirical measure to be used to proxy for such fac- tors/mechanisms” (Carmignani et al., 2014, p. 516). For example, should we use GDP per capita (as four studies considered in this paper do) as a proxy, the log of GDP per capita (two studies), or the log of GDP at purchaser’s price parity (two studies)? Should we use one measure or two? How would results change, for example, if instead of using Laakso and Taagepera’s (1979) effective number of parties, one decides to use the dis- aggregated number of parties or the number of parties weighted by their vote share? Why Extreme Bounds Analysis? By contrast, one of extreme bounds analysis’ key characteristics is that it estimates all possible combinations of a set of predictors showing how slight changes in the included variables affect estimation results. Therefore, rather than focusing on a specific set of variables, by considering all possible variable combinations, EBA can suggest which predictors are systematically robust. What EBA cannot do as well as meta-analysis is theoretically and empirically highlight the relationship between a particular out- come and explanatory variable, including possible mediation and moderation by other factors. An extreme bounds analysis estimates a set of regressions with the following func- tional form: =+++, 40 613 1 3 Political Behavior (2023) 45:607–633 where Y is the dependent variable (in our case, voter turnout), I is a vector of core variables included in all models, M is the variable of interest, Z is a set of controls, and u is the error term (Levine & Renelt, 1992). I is the set of variables always included—the “base” or “core” variables—because the literature suggests a well- established relationship with the dependent variable. What is the I vector of vari- ables for electoral turnout? Unfortunately, as we show below, less than those that one, a priori, may think. EBA repeatedly estimates the equation with a different set of Z controls in each regression. Since every regression produces a coefficient for all included variables, all the regressions create a distribution of such coefficients. In order to decide whether the coefficients are robust, researchers have relied on two main criteria. The first by Leamer (1983) suggests that a variable should be considered robustly related to the outcome variable if, and only if, the lower and upper extremes of a variable’s coefficients have the same sign.10 Specifically, the extreme upper (lower) bound is defined by the maximum (minimum) value of the variable of interest plus (minus) two standard deviations. If the variable of interest remains of the same sign at both upper and lower bounds, then such relation among the variables is said to be “robust.” When the variable of interest does not keep the same sign at both upper and lower bounds, then such relation among the variables is said to be “fragile.” In short, only after running all possible regressions including all variables and only if all estimates are in the same direction, are results considered robust. Sala-i-Martin (1997) finds Leamer’s standard to be overly restrictive in most cases because it is likely that if enough model specifications are analyzed, and assuming that the dis- tribution of β has both some positive and negative support, it is likely that the signs of the coefficients will change at least once. In fact, following Leamer’s criteria if a single regression produces a coefficient of the opposite sign large enough to shift one of the bounds, then the variable is considered not robust. Sala-i-Martin (1997) proposes to look at the entire distribution of coefficients and conclude a variable is robustly related to the outcome variable when a large percentage—say 90–95%—of the coefficient’s distribution is either above or below zero. These criteria, then, can lead to different substantive conclusions. In the extreme, if enough regressions are run and the distribution of the estimators have some “posi- tive and some negative support, then one is bound to find one regression for which the estimated coefficient changes signs” (Sala-i-Martin, 1997, p. 179). Indeed, by following Leamer’s approach, we may conclude that the knowledge about many social phenomena is scarce and thereby make a consequential Type-II error. In con- trast, as Plümper and Traunmüller (2020, p. 149) recently show, Leamer’s EBA has “an extremely low probability of producing false positives” while, Sala-i-Martin is “more likely to suffer from identifying false positives than the inferential rule it replaced” (Plümper & Traunmüller, 2020, p. 149). Previous works have, in the main, relied on Sala-i-Martin’s rather than Leamer’s approach.11 However, both criteria 10 A criterion used by Levine and Renelt (1992). 11 Sala-i-Martin’s (1997) focus on the entire coefficient distribution is a common approach used by Hegre and Sambanis (2006), Gassebner, Lamla and Vreeland (2013), Hartwig and Sturm (2014), Gassebner, Gutmann and Voigt (2016), and Miller, Joseph and Ohl (2018). 41 614 Political Behavior (2023) 45:607–633 1 3 are informative, so we follow Hafner-Burton (2005) and report results according to both criteria. As is clear below, there are substantive differences in what is consid- ered robust predictors of turnout using these different criteria, and we compare our results using both criteria. Research Design In the turnout literature, there has been a thorough discussion of how to best opera- tionalize the phenomenon of interest. The two main options are the number of votes cast as a percentage of the voting age population (VAP) or the voting registered population (VRP). There are arguments for both. For example, Blais and Dobrzyn- ska (1998) use the latter and argue that VAP is not adjusted for the alien population, which artificially downplays turnout. Endersby and Krieckhaus (2008, p. 602), by contrast, recommend VAP because if registration is not automatic, and registration and voting are correlated, “then the ratio of voters to registered voters is a biased measure of citizen’s motivation to vote.” Among our 44 studies, 16 use VAP, 18 use VRP, 5 use both, 4 use other definitions12 and 1 (Siaroff & Merer, 2002) does not provide a definition. In this article, we primarily use VRP, but we also use VAP measures in a series of robustness checks.13 If after applying the same analysis on two related but different dependent variables, the results of the independent vari- ables are similar, this would be a clear signal of a variable’s strength. Turnout data are from International IDEA (2017). According to IDEA, their data comes from the national election management bodies (EMBs) and national statistical bureaus. EMBs provide data from their official reports and web portals. IDEA’s population data comes from secondary sources. In order to be included in the dataset, the elec- tion has been held after 1945; must have been for national political office in an inde- pendent nation state; there must more than one party contesting the election; and the franchise must be universal.14 In our data, the VAP and VRP turnout measures correlate at 0.68. Consistent with the literature we limit our sample to lower house elections in democracies (defined as a Polity value of six and above in the year before the observed election). Overall, our sample includes 579 elections in 80 democratic countries from 1945 to 2014. 12 Such definitions measure turnout as “the total votes cast divided by the size of the electorate” (Blais & Carty, 1990, p. 169); “the average turnout of each country” (Colomer, 1991, p. 319); the proportion of the eligible electorate voting (Radcliff, 1992, p. 445); and “the percentage of eligible voters that turned out at the respective country’s national election” (Stockemer, 2015, p. 87).13 This debate is not new. In the United States, McDonald and Popkin (2001) proposed another measure, but no cross-national data for this measure exists. More recently, Stockemer (2016) created VEP for 500 elections in 116 countries. Given our focus here, we rely on the established measures. 14 Despite all the efforts, IDEA’s data are not perfect. For example, when two elections were held in a single year, IDEA does not report which election is captured. We thank a reviewer for highlighting this fact. 42 615 1 3 Political Behavior (2023) 45:607–633 Independent Variables In order to identify the most common predictors of voter turnout, we rely on Geys (2006), Geys and Cancela (2016), and Stockemer (2017) meta-analyses of 83, 185, and 130 works, respectively. For our analysis, we included all English language, national-level, comparative, peer-reviewed journal articles focused on voter turnout published between 1980 and 2017. We therefore exclude case studies, studies focus- ing on local, regional, or provincial elections, book-length studies, and studies not in English. We focus on the national level because local dynamics are likely dis- tinct from those at the national level. Furthermore, logistically it also makes sense to exclude works where the underlying data are not comparable to other cases: for example, exploring the effects of Norwegian school referendums (Kaniovski & Miller, 2006) on voter turnout in non-Norwegian countries is not possible. This decision implies that some factors, like campaign expenditures, cannot be examined given the almost total lack of available data outside the US. The 44 included studies are listed in the Appendix.15 As mentioned above, turnout predictors are usually organized into three groups: socio-economic, institutional, and political; and we follow this approach when organizing 127 independent variables derived from the 42 articles we examine.16 We find 41 socio-economic factors, 48 institutional factors, and 38 political factors. Such a large number of independent variables reinforce the diversity of empirical approaches in the literature and the need to clearly determine what robustly affects turnout (and what does not). Tables 1, 2 and 3 summarizes each group of variables. The first column presents the number of times a variable is used in the literature; the second column includes the variable name or concept; the third column show the ways in which the variable has been measured (if available); the fourth column present the directional effect in turnout—sub-divided in four sub-columns, one accounting for each possible result. When the variable had a positive impact for turnout, it is labelled as ‘positive’, ‘neg- ative’ when the contrary; ‘NS’ suggests a non-significant result, while ‘mixed’ cap- tures those results when the results vary depending on the model. Table 1 includes 48 institutional factors. The most frequent factor is compulsory voting, which is measured in three different ways. Twenty-eight studies find that compulsory voting has a positive and significant effect while five find it not signifi- cant. Two other variables—the number of political parties and proportional repre- sentation (PR)—are the next frequent (19 times each). This illustrates the literature’s differences in measurement; the former is operationalized in 10 different ways, the latter in two. We have created a straightforward measure of agreement among stud- ies by dividing the most frequent result by the number of studies that use such meas- ure and multiplying it by 100. The higher the percentage, the more established the 15 Appendix Table A1 lists the 285 studies that were excluded and the reasons for exclusion. 16 For example, see Geys (2006) and Blais and Drobzinska (1998). 43 616 Political Behavior (2023) 45:607–633 1 3 finding.17 For compulsory voting, the degree of agreement is over 84%. Other vari- ables are less established. For example, the agreement on the impact of the number of parties or the impact of proportional representation electoral systems is unclear, with about 53% agreement. Regarding the former, while 10 cases find that higher number of parties, lead to lower turnout, 8 do not find it significant. Likewise, 10 studies show that countries under proportional representation systems show higher turnout while 7 find it not significant. Table 2’s socio-economic variables include two sub-groups, socioeconomic characteristics and geographical dummies. First, the indicators gathering relevant socio-economic characteristics of a given society, such as size of population, GDP, GDP growth, and urbanization. Twenty-two studies include GDP as an independent variable. Eight find that GDP has a positive impact on turnout, while seven show a non-significant relationship, three report a negative influence, and four show mixed results. Results are not much better for the rest of such variables. Second, there are geographical dummies for specific countries or regions. Most notable about these variables is that including a variable for Switzer- land and/or for the US almost always are negatively related to turnout. There are two patterns worth mentioning when discussing Table 3’s political variables. First, not many political variables appear in the articles we examined. This is surprising given the fundamentally political nature of turning out to vote. An exception is “closeness/ competitiveness” that appears in 21 studies (almost half of our sample). Further- more, the level of agreement for this variable is below 50%. Second, we only find a high level of agreement for the previous election turnout level (though such variable only appears in five studies). In sum, three important findings arise from this initial literature review. First, out of 127 distinct variables, less than half (44%) appear more than once. Even the most frequently used indicator, compulsory voting, was included in less than 75% of the examined studies. Second, among the 55 variables that appear in more than 1 article, over half (57%) are measured in more than one way. Third, it seems more generally that turnout studies face a paradox—while voting is mainly a political act, the most common empirically tested arguments in the literature are of institutional or socio- economic mechanisms. Only recently have some articles examined the impact of terrorist attacks (Robbins et al., 2013), corruption (Stockemer et al., 2013) or elec- toral dynamics (Martínez i Coma & Trinh, 2017) on turnout. Table 4 condenses this information and also offers an overview of the distribu- tion of the 70 variables for which we have data. These results strongly suggest that a standard model of turnout does not yet exist, and few factors—especially institu- tional and socioeconomic—have a consistently established effect on voter turnout. 17 This points us to a degree of agreement about a specific covariate. We define 70% or more as a “high level of agreement.” When comparing the common variables from Geys (2006) and Stockemer (2017), only compulsory voting shows a high level of agreement (84%), followed by income inequality (60%), PR (53%), vote closeness (47%) and population size (44%). 44 617 1 3 Political Behavior (2023) 45:607–633 Table 1 Institutional predictors of voter turnout Frequency Indicator # of ways of measure- ment Results % Agree Positive Negative NS Mixed 33 Compulsory voting 3 28 5 84 19 # Parties 10 10 8 1 53 19 Proportional representation 2 10 1 7 1 53 18 Concurrent/Simultaneous 2 13 4 1 72 13 Unicameralism 3 4 1 7 1 54 9 Disproportionality 6 3 6 67 9 District Magnitude 5 3 1 5 55 8 Automatic (voluntary) registra-tion 2 3 1 4 50 8 Legal voting age 3 1 4 3 50 5 Federalism 3 1 4 80 4 Age percentages 3 1 1 2 50 4 Plurality 3 1 2 1 50 4 Majority 1 3 1 75 4 Female suffrage 3 2 1 1 50 4 Mixed/semi-presidential system 1 4 100 2 Leg. party/parl. fractionaliza-tion 2 1 1 50 2 One-party majority govern- ment 1 2 100 2 New voter expansion 1 2 100 2 Effective electoral threshold (ln)1 2 100 2 Cumulative executive respon-siveness 1 1 100 2 Cumulative absentee ballots 1 2 100 1 Parliamentary system 1 1 1 Relevant elected president 1 1 1 Voting holiday 1 1 1 Size of legislature (ln)1 1 1 Majority (multi-member)1 1 1 # of parties, squared 1 1 1 % Market of state-owned enterprises 1 1 1 Private broadcast system 1 1 1 Mixed system (definition #1)1 1 1 Mixed system (definition #2)1 1 1 Partisan press 1 1 1 Public broadcasting audience 1 1 1 Newspaper subscriptions 1 1 Campaign funding limits 1 1 45 618 Political Behavior (2023) 45:607–633 1 3 Estimation Considerations Including all 127 predictors discussed above in our empirical models is not possible due to data availability—several variables are available for only a short span of time, specific region (i.e., Europe) or a clear set of countries (i.e., OECD members).18 The final line in Table 4 describes the distribution of the seventy variables we include, and Appendix Tables A2, A3 and A4 present these variables’ summary statistics, operationalization, and sources. Given the absence of a commonly accepted model of voter turnout, in our selec- tion of core variables we rely on several theoretical assumptions consistent with the literature and empirical regularities to include five variables in all of our models. From an institutional perspective, compulsory voting has been found to affect turn- out. 84% of our 44 studies find that compulsory voting increases turnout, which is likely the literature’s most commonly accepted finding. The electoral system is also a recurrent variable of study under the (challenged) assumption that voter turnout is usually higher in proportional representation systems (Blais & Carty, 1990). We also include two socio-economic factors—per capita gross national income and popula- tion (both logged to control for outliers). The former accounts for the literature’s finding that economic development fosters turnout while the latter controls for the NS not significant Table 1 (continued) Frequency Indicator # of ways of measure-ment Results % Agree Positive Negative NS Mixed 1 Public direct funding 1 1 1 Free TV access 1 1 1 Access to paid TV ads 1 1 1 Cohabitation 1 1 1 % Legislators elected in national districts 1 1 1 Proportional representation seats (%)1 1 1 Strong regional governments dummy 1 1 1 Party membership 1 1 1 Polarized party system dummy 1 1 1 Personal vote 1 1 1 Direct election 1 1 1 Compulsory voting enforced 1 1 1 Cumulative female empower- ment 1 1 18 Over 80% of excluded variables appear in only one study. 46 619 1 3 Political Behavior (2023) 45:607–633 Table 2 Socio-economic predictors of voter turnout NS not significant Frequency Indicator # of ways of measurement Results % Agree Positive Negative NS Mixed 22 GDP/GNI 14 8 3 7 4 36 16 Population size 3 1 7 7 1 44 10 Switzerland dummy 2 10 100 9 GDP/GNI growth 8 1 1 7 78 8 Urbanization 2 3 4 1 50 7 Literacy rate/illiteracy/education/secondary school enrolment 5 1 4 2 57 5 Inequality 3 2 3 60 5 USA dummy 1 4 1 80 3 Population density 1 1 1 1 33 3 Average life expectancy 2 3 100 3 Public expenditure 2 1 1 1 33 3 Ethnic fractionalization 1 2 1 67 3 Electorate size 2 1 2 67 2 Latin America dummy 1 2 100 2 Africa dummy 1 1 1 50 2 Asia dummy 1 2 100 2 Oceania dummy 1 2 100 2 West dummy 1 2 100 2 Socioeconomic development/Human Development Index 2 1 1 50 2 Ethno-ling. fractionalization 2 1 1 50 2 Inflation 2 1 1 50 2 Unemployment 1 2 100 2 Unionization 2 2 100 2 Corruption 2 2 100 1 GNP growth 1 1 1 South America 1 1 1 Authority/Decentralization 1 1 1 Revenue/Decentralization 1 1 1 Spending/Decentralization 1 1 1 Sweden 1 1 1 New Zealand 1 1 1 Norway 1 1 1 OECD 1 1 1 Media exposure 1 1 1 Union density 1 1 1 KOF economic globalization 1 1 1 Islamic majority nation 1 1 1 Eastern Europe 1 1 1 Linguistic fragmentation 1 1 1 Religious fragmentation 1 1 1 Major religions 1 1 47 620 Political Behavior (2023) 45:607–633 1 3 relationship between community size (an element assessed both by Geys, 2006 and Stockemer, 2017) and turnout. Finally, we include the level of turnout in the previ- ous election because the literature suggests that turnout may be habit forming and has its own inertia (Geys, 2006).19 While we acknowledge that more core variables could be included, we believe our core turnout predictors are consistent with the literature we survey. Given that several variables measure similar concepts and/or that the measure- ment of such variables may be related, multicollinearity may be a risk when includ- ing such a large number of predictors. Levine and Renelt (1992, p. 945)20 rely on three common strategies to reduce this risk: (1) they limit the total number of explanatory variables to eight at most; (2) they limit the number of Z controls to seven; (3) and they further restrict their Z variables by excluding variables that a pri- ori may measure the same phenomenon. Similarly, Hartwig and Sturm (2014) drop some variables when they are highly correlated. To avoid artificially inflating our estimates, we have followed their first and third strategies (we limit the total number of core and independent variables to eight and we exclude multiple measures of the same underlying phenomena).2122 The final estimation consideration is what type of model to run. The most com- monly used method in the 44 reviewed studies was ordinary least squares (OLS) with robust standard errors clustered by country (15 articles). Regression models with fixed effects and random effects were used in several (6 articles and 7 arti- cles respectively) of the 44 articles examined in the literature review above.23 Nev- ertheless, there are reasons to expect with eight predictors that there is the real risk of unobserved variable bias and risk of unmeasured unit-effects. Furthermore, like most of the literature we are interested in the reasons for differences in turnout both between and within countries. There is an ongoing methodological debate about which is appropriate for time-series cross-section data such as ours that is outside the scope of this article.24 Therefore, we estimate two series of models. The first series of OLS models with fixed effects focuses on time-varying explanations for turnout. Our I vector includes 5 variables, and the Z vector has 53 variables. The second series of models use random effects and clustered standard errors by country to include time-invariant variables. A random effects approach relies on the strong assumption that the included predictors are uncorrelated with country-specific inter- cepts (i.e., unobserved factors), but it does allow for time-invariant factors’ effects 19 We are aware of Achen’s (2001) warning regarding the inclusion of a lagged dependent variable. Our further analyses below address this concern.20 This article examines economic growth and over 50 independent variables. 21 The third strategy to avoid multicollinearity is included in our design by default because we rely only on the most frequent source and variable operationalization.22 We also lag time-varying independent variables to reduce the risks of endogeneity or reverse causal- ity.23 Models took an average of 12.5 days to run 2,433,115 regressions using Stata 16.1 on an Amazon Web Service Elastic Compute Cloud c3.large instance running 2019 Windows Server Edition.24 See for example Bell and Jones (2015), Bell, Fairbrother and Jones (2019), Clark and Linzer (2015), Dieleman and Templin (2014), Imai and Kim (2019), and McNeish and Kelley (2019). 48 621 1 3 Political Behavior (2023) 45:607–633 Table 3 Political predictors of voter turnout NS not significant Frequency Indicator # of ways of measurement Results % Agree Positive Negative NS Mixed 21 Closeness/competitiveness 5 3 9 9 45 11 Democracy 9 5 1 5 45 5 Previous turnout 1 4 1 80 4 Founding elections 1 3 1 75 3 Number of elections 3 1 2 67 2 New democracy 1 2 100 2 Third election 1 2 100 2 Fourth election 1 2 100 2 Short term majority status of gov’t 1 2 2 Short-term margin of victory 1 2 2 Short-term cohesiveness 1 2 1 Recent threat to democracy 1 1 1 Second election 1 1 1 Third or fourth election 1 1 1 Fifth or sixth election 1 1 1 Electoral volatility 1 1 1 Voting share of left parties 1 1 1 Frequent changes in executive 1 1 1 Average strength of party group linkages 1 1 1 Party polarization 1 1 1 Post-materialist party in parlia-ment 1 1 1 Prior democratic experience 1 1 1 Terrorist attack 1 1 1 Party replacement 1 1 1 Two-party vote concentration 1 1 1 Seat ratio—first to second parties 1 1 1 Number pre-electoral coalitions 1 1 1 Dispersion 1 1 1 Opposition harassment 1 1 1 Opposition ban 1 1 1 Boycott 1 1 1 Violence 1 1 1 Party linkages 1 1 1 Seventh or eighth election 1 1 1 Decisiveness 1 1 1 Short-term mean margin 1 1 1 Authoritarian path 1 1 1 Democratization path 1 1 49 622 Political Behavior (2023) 45:607–633 1 3 on turnout to be estimated. Our I vector for the random effects models includes 5 variables, and the Z vector has 65 variables. In both series of models, continuous independent variables are standardized in order to aid the comparability of coeffi- cients and a number of continuous predictors were log-transformed to reduce the effects of extreme values.25 Results After running 1,170,324 regressions with fixed effects, the 5 core variables, and 22,096 unique combinations of 53 independent variables, nine variables are robust according to Sala-i-Martin’s criteria [the area under the general cumulative density function (CDF) that are on either side of zero is less than 0.05].26 One variable (eco- nomic globalization) also meets Leamer’s (1985) more demanding criteria. Table 5 summarizes the results for the variables our fixed effects analysis suggests are robustly associated with voter turnout.27 The nine robust predictors are evenly split between institutional, socio-economic, and political factors. The three institutional variables are proportional representation, concurrent elections, and the number of years since universal suffrage was introduced—the first two increase turnout, while the last decreases turnout. As for socioeconomic conditions, economic globalization decreases turnout and inflation, and spending decentralization increases it. Among the three political factors (competitiveness, a dummy for elections before 1995, and a time trend variable), the most interesting result is that higher levels of competitive- ness (measured as the difference between the first and the second party’s vote share) decreases turnout. Only one (proportional representation) of the core five variables Table 4 Variable overview Frequencies in parentheses Socio-economic Institutional Political Total Variables 41 48 38 127 % of common variables 58 (24)44 (21)29 (11)44 (55) % of common variables measured differently 62 (15)66 (14)27 (3)57 (32) Note: Absolute figures in parentheses  Variables mentioned 41 48 38 127  Variables included in EBA models 33 16 21 70 25 There are two possible sources of variation: (1) substantively important results to small changes in the variables included and (2) how fragile the inferences are to small changes in operationalization. Given that some variables appear nineteen times but are operationalized in ten different ways, we focus on the first source of variation. Even in this case, 70 variables in the random effects models lead to over 2 mil- lion regressions. We leave for future research the analysis of different operationalizations.26 Sala-i-Martin (1997, pp. 179–180) looks at both normal and generalized CDFs. Histograms of the 1.2 million estimated coefficients (available on request) suggest most coefficients are not normally distrib-uted, so we use the generalized CDF. 27 Table A7 summarizes results for all 58 variables. 50 623 1 3 Political Behavior (2023) 45:607–633 was found to be robust. Finally, it is notable that several variables our analysis finds significant have been largely overlooked by the existing literature. For instance, eco- nomic globalization and spending decentralization were each included in only 1 of the 44 articles we analyze, yet they are both robust predictors in these and other models discussed below. Our fixed effects model results suggest that 9 out of the 53 time-varying variables we measure are robustly associated with national-level turnout. However, as men- tioned above these models exclude 12 time-invariant or sluggishly changing vari- ables that previous research has found important. Therefore, we run an additional 2.9 million regressions with random effects that allow us to capture important vari- ation in turnout across countries and regions. Our findings (summarized in Table 6) suggest that seven variables meet Leamer’s robustness criterion and 14 variables meet Sala-i-Martin’s.28 Seven are location dummies—Eastern Europe, Latin Amer- ica, and Switzerland dummies depress national-level voter turnout while Norway, New Zealand, Oceania, and Sweden dummies have the opposite effect. From the socio-economic indicators, higher ethnic fractionalization and levels of inequality are robustly associated with lower electoral turnout while inflation increases turnout. Additionally, the lagged dependent variable is a robust turnout predictor (unlike in the fixed effects models). Finally, two institutional factors (concurrent elections and compulsory voting) are also robustly associated with higher turnout. The results for these two institutional variables (concurrent elections and com- pulsory voting) are also interesting because they highlight the different substantive conclusions that would be reached depending on which robustness threshold is cho- sen. Both variables are significant using Sala-i-Martin’s criterion but not Leamer’s. Let us explore for a bit why this is the case. In almost three million regressions using random effects, the concurrent election variable never has an estimated coefficient below zero (the smallest is 0.04). This is why this variable’s CDF never includes negative values and therefore why Sala-i-Martin would consider this variable a robust predictor of turnout. At the same time, because of concurrent election’s aver- age standard deviation (7.01) is similar in size to its average coefficient (7.89), its lower extreme bound (two standard deviations below the mean) is less than zero and therefore not robust for Leamer. The story behind compulsory voting’s results is different because (out of 2.9 million random effects models) 9775 models (0.003%) do indeed produce a negative coefficient for this variable. Is there something sys- tematically different about these 9775 models that lead to such a counter-intuitive result? Several things stand out.29 First, the inclusion of several variables substan- tially reduce the sample size. Union density is in 9430 (96%) of these 9775 mod- els, followed by human development index (6616, 68%) and the Oceania dummy (1500, 15%). Second, democracies with union density data are relatively rare (17 of 80) leading to the possibility that outliers can have undue effects on estimated coefficients. This might help explain why the average number of observations (40) in these 9775 models are significantly lower than for the other 2.9 million models 28 Complete results available in Table A8. 29 Table A15 includes the variables included in these 9775 models. 51 624 Political Behavior (2023) 45:607–633 1 3 including compulsory voting (360).30 Figure 1 provides a striking visual of this dif- ference in sample size. It is worth noting that (as in all articles using EBA that we are aware of) our regressions include varying sample sizes due to data availabil- ity. This is one of EBA’s main strengths—if a variable significantly affects turnout across a range of samples, this is a clear signal of the robustness of the estimated relationship. Turning now to other ways our results are consistent (or not) with the previous work summarized in Tables 1, 2 and 3, we reach several main conclusions. First, the mixed results for two of our core variables (population and GNI) is consistent with the literature. Take population. Table 2 suggests that population size has an incon- clusive effect on turnout—seven studies find a negative effect and another seven find no significant effect. Our fixed effects and random effects results also suggest that population and economic development have no robust effect on turnout. Simi- larly, we do not find significant effects for several common socioeconomic variables including economic growth (nine studies) and literacy (seven studies). Regarding institutional variables, we do not find robust results for the number of parties, the different electoral formulae, the magnitude of the electoral district, the level of dis- proportionality, and the legal voting age, among others.31 By contrast, our results suggest that economic globalization (one study) and inflation (two studies) affect turnout while receiving nowhere near as much attention as a number of institutional Table 5 Fixed-effects extreme bounds analysis of voter turnout Ave β the average coefficient value, SE standard error, % Sign. percentage of models with a statistically significant (p < 0.05) coefficient, CDF cumulative density function below 0 *Variable significant using Leamer’s criteria. Complete results reported in Table A7 Variable Models Ave. β Ave. SE % Sign CDF < 0 Competitiveness 22,096 − 1.64 1.16 15.62 0.975 Concurrent 22,048 10.15 6.45 31.14 0.007 Economic globalization*22,096 − 4.27 1.82 75.11 0.998 Inflation 22,096 0.89 0.68 15.61 0.029 Spending decentralization 22,096 4.77 8.89 10.16 0.032 Suffrage 21,987 − 2.02 1.27 15.81 0.964 Time trend 22,096 − 6.87 8.87 61.25 0.987 Years 1945–1994 22,096 1.73 0.98 12.02 0.043 Core model Compulsory voting 1,038,770 8.62 7.96 27.91 0.160 GNI per capita, ln 1,170,324 − 12.02 11.77 37.58 0.908 Lagged dep. var 1,170,306 − 0.06 0.27 14.03 0.429 Population, ln 1,169,987 2.45 1823.88 3.75 0.658 Proportional representation 1,128,068 5.18 3.31 33.90 0.026 30 Negative coefficient model average sample size: 40 (s.d. 33.8); all models: 360 (s.d. 179.3). 31 It is important to note that all of these variables also had mixed results in the literature. 52 625 1 3 Political Behavior (2023) 45:607–633 Table 6 Random-effects extreme bounds analysis of voter turnout Ave β the average coefficient value, SE standard error, % Sign. percentage of models with a statistically significant (p < 0.05) coefficient, CDF cumulative density function below 0 *Variable significant using Leamer’s criteria. Complete results reported in Table A8 Variable Models Ave. β Ave. SE % Sign CDF < 0 Concurrent 41,600 7.88 7.01 23.63 0.000 E. Europe*34,220 − 4.46 1.43 64.06 0.992 2nd election*35,750 − 6.11 1.35 82.69 0.998 Ethnic fractionalization 41,660 − 1.89 12.64 16.85 0.987 GINI index 41,537 − 4.24 7.85 35.26 0.989 Inflation 41,658 0.96 0.45 12.70 0.010 Latin Am. and Caribbean 34,220 − 4.84 1.90 62.24 0.992 Norway*41,303 2.10 1.33 23.23 0.035 New Zealand*35,990 5.06 2.10 73.25 0.041 Oceania 41,599 6.55 4.47 80.89 0.002 Sweden*41,303 5.41 1.70 74.25 0.003 Switzerland*41,231 − 12.41 4.44 90.08 1.000 Core model Compulsory voting 2,955,366 4.87 9.61 63.95 0.003 GNI per capita, ln 2,956,512 − 0.65 6.05 20.20 0.311 Lagged dep. var.*2,956,512 0.63 0.16 92.50 0.008 Population, ln 2,956,017 − 1.47 6.40 4.85 0.928 Proportional representation 2,936,444 0.80 1.92 2.86 0.246 0 2,000 6,000 4,000 Frequency0 50 100 150 200 250 300 350 400 450 500 550 600 0 50000 100000 150000 200000 Frequency0 50 100 150 200 250 300 350 400 450 500 550 600 Number of observations Fig. 1 Sample sizes for 384 models with negative coefficients for compulsory voting (top) and in all (2,955,366) random effects models with compulsory voting (bottom) 53 626 Political Behavior (2023) 45:607–633 1 3 and political factors (e.g., 19 articles include the number of political parties in their models). It is also notable that (regardless of the criteria we rely upon and exclud- ing geographical dummies) the most robust turnout predictors are political while (as we saw above) previous research focuses more on institutional and socioeconomic factors. Finally, it is worth highlighting our models’ average substantive effects. Figure 2 summarizes the coefficient distributions for Table 5’s robust predictors. Each square represents a variable’s average estimated impact on national-level voter turnout, and the bars on either side of these coefficients represent Leamer’s upper and lower extreme bounds. Holding all else equal, concurrent elections increase turnout 7.9% over elections without candidates competing for executive office. Proportional rep- resentative election systems are associated with higher levels (5.2%) of voter turnout than states without this electoral system. The variables that (on average) depressed turnout the most in the fixed effects models are the time trend (which estimated a lowering of average turnout by 6.9% from 1945 to 2014) and economic globaliza- tion (− 4.3% average effect across its observed range). Time trend Econ. globaliz. Sufferage Closeness/compet. Inflation Years 1945-94 Spending decent. Concurrent 1.5 Proport. represent. * GNI pc, const. (ln) * Compulsory voting * Lagged D.V. * -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 Coefficient Fig. 2 Robust predictors of voter turnout, mean coefficients and extreme bounds. Note Results from fixed effects models reported in Table 5. Table A7 includes complete results. *Identifies core variables in all fixed effects models. The only core variable found to be (Sala-i-Martin) robust is proportional repre- sentation. Population is a core variable but not included in figure due to the disproportionate size of its extreme bounds 54 627 1 3 Political Behavior (2023) 45:607–633 Further Analyses Our empirical results taken from over four million regressions using fixed and ran-dom effects provide compelling evidence for the robustness of a number of predic-tors of national-level voter turnout. These results, however, vary more in the vari-ables the models included than their model specifications themselves, which lead to several potential limitations. First, we run our main models using fixed and random effects rather than the literature’s most common model specification—OLS regres-sion using robust standard errors clustered by country. Second, they include a lagged dependent variable, which some (e.g., Achen, 2001) argue depress predictors’ sta-tistical significance. Third, we focus on registered voter turnout, while the literature also suggests turnout as a percentage of voting age population could lead to differ-ent results. Fourth, these seventy factors may work differently in different types of countries (e.g., developed/developing, Western/non-Western), and fifth there may be over-time variation that is not captured by our time trend variables. We, therefore, re-run all or part of our analyses with fourteen additional model or sample varia-tions. Figure 3 summarizes the results from these additional 11.8 million models for the variables included in Tables 5 and 6.32 Using robust standard errors First, we run 2,433,115 models using robust clus-tered standard errors the core 5 variables, and 65 alternating M and Z variables. 11 and 20 variables were significant using either Leamer or Sala-i-Martin’s criteria. Fourteen of these 20 (70%) variables were also significant in either the fixed effects or random effects models. Six variables (boycott, a fifth or sixth election dummy, GNI growth, quality of democracy, union density, and unionization) were robust in these new models but not in our main models, although four of these six (fifth or sixth election, quality of democracy, union density, and unionization) would be considered robust in either the fixed or random effects models if we use a 0.1 level rather than 0.95. Thus, our first set of further analyses suggest that the majority of our results using fixed effects and random effects hold if we use robust standard errors clustered by country.33 Removing the lagged dependent variable Next, we run an additional 4,228,480 regressions identical to the three series of models described above except that we lagged turnout. We included it above because the literature suggests today’s turnout is systematically related to previous turnout. However, as Achen (2001) explains, lagged dependent variables may suppress the explanatory power of relatively time-invariant independent variables, such as compulsory voting or population size. The results for models without a lagged dependent variable do vary in a few ways, a few variables lose robustness, while more than double are now found to be robust. Spe-cifically, in the fixed effects models without a lagged dependent variable, concurrent elections now meet the Leamer robustness criteria while suffrage is no longer signif-icant at either level. Otherwise, the other 54 variables in these fixed effects models 32 Appendix Tables A7 to A18 includes more detailed results for these analyses. 33 70% of variables in fixed effects models and 83% of random effects variables. 55 628 Political Behavior (2023) 45:607–633 1 3 (96%) have substantively identical results. Turning to the random effects models, economic globalization, the time trend, the 1945–1994 dummy, and competitive- ness are now significant and in the same direction as the fixed effects models. Our population variable is also now significant, while the second election dummy loses significance. Overall, 57 of 69 variables (83%) have substantively identical results. Finally, in the models with robust standard errors and no lagged dependent vari- able concurrent elections and the second election loses significance without lagged dependent variable while population gains significance. Here, 58 of 69 variables (84%) have substantively similar results. Overall, these three series of models with- out lagged dependent variables produce quite substantively similar results as those models with a lagged term. Changing denominators Next, we re-estimate the lagged dependent variables and clustered standard errors models using a different dependent variable denomi- nator (voting aged population instead of registered voters). The literature summa- rized above splits almost evenly in its use of these denominators. Given that these variables’ correlation coefficient is 0.67, results may significantly vary depending on which denominator used. Overall, we find 51 of 70 (73%) variables’ robustness unchanged. Twelve of seventy variables are robust according to both criteria. Four are geographical dummies (Sweden, Switzerland, Easter Europe, and Latin Amer- ica/Caribbean), and the others are now familiar (compulsory voting, concurrent election, ethnic fractionalization, inflation, previous turnout, second election, fifth or sixth election, time trend, and union density). Eight variables are no longer robust (boycott, GNI growth, the Latin American and Oceania dummies, quality of democ- racy, spending decentralization, unionization, and years 1945–1994), while 11 are now considered robust (the Asia, Norway, and USA dummies, economic globaliza- tion, female suffrage, plurality system, radios per capita, third election dummy, leg- islative seats, share of voters aged 30 to 69, and violence) in this series of over 2.3 million models. Splitting samples Different samples of countries may also have systematically different correlates of national-level voter turnout (Stockemer, 2015). Therefore, we ran an additional 2,647,887 analyses breaking up our sample into different groups—democracies and all states, Western and non-Western states, established and newer democracies, and above and below median income states.34 Our results (see Appendix) suggest that while we limit our main analysis to democratic states, turnout drivers are overwhelmingly similar in non-democracies.35 Indeed, 64 of 70 variables’ robustness does not substantively change in models including anocracies and autocracies (i.e., states with a Polity score of less than six). Of the six variables with changes, four are robust in all-states models and not in democracy-only models (Asia, revenue decentralization, urban population, and violence), while two are no longer robust (boycott, and the 1945–1994 dummy). The vast majority (12 of 15) of variables in Western and non-Western democracies did not significantly change 34 States that had Polity2 scores of six or greater for 20 years or more are considered “established”.35 Turnout in non-democracies is rarely studied in a comparative framework. Exceptions include Mar-tínez i Coma (2016) and Martínez i Coma and Morgenbesser (2020). 56 629 1 3 Political Behavior (2023) 45:607–633 from the original models. For those that did change, boycotts, per capita GNI, and spending decentralization were robustly associated with turnout in Western states but not in non-Western states. The results for the latter are likely to be at least partly driven by increased data availability in Western states. The most notable differ- ences in these series of split sample models were found between countries that had more or less than two decades since democratization. Take, for example, the East- ern Europe dummy that is robust in the main models, it robustly decreases expected turnout in established democracies but not in the new democracies. Relatedly, once the analysis is performed in newer democracies, regional differences fade away— a result consistent with previous studies (e.g., Kostadinova & Powers, 2007).36 Additionally, the original findings hold for a number of variables by countries with above the median income than below it. Of the 22 robust predictors described above, the richer country models had 9 significant predictors while the poor states had 4. VTA VAP VTNL (l, s) VT (l, s) VTNLRE (l, s) VTNLFE (s) VTRE VTFE (s) 0 5 10 Years 1945-1994 VTA VAP VTNL VT VTNLRE (s) VTNLFE (s) VTRE VTFE (s) -10 -5 0 5 Closeness/competitiveness VTA (s) VAP (l, s) VTNL (l, s) VT (s) VTNLRE (l, s) VTNLFE VTRE (s) VTFE -20 -10 0 10 20 30 Compulsory voting VTA (s) VAP (s) VTNL VT (s) VTNLRE (s) VTNLFE (l, s) VTRE (s) VTFE (s) -10 0 10 20 30 Concurrent elections VTA VAP (s) VTNL VT VTNLRE (l, s) VTNLFE (l, s) VTRE VTFE (l, s) -8 -6 -4 -2 0 2 Economic globalization VTA VAP VTNL VT VTNLRE VTNLFE VTRE VTFE -40 -20 0 20 GNI pc, ln VTA (s) VAP (l, s) VTNL (l, s) VT (s) VTNLRE (l, s) VTNLFE (s) VTRE (l, s) VTFE (s) -1 0 1 2 3 4 Inflation VTA (s) VAP (l, s) VTNL VT (l, s) VTNLRE VTNLFE VTRE (l, s) VTFE -.5 0 .5 1 Lagged dependent var. VTA VAP VTNL (s) VT VTNLRE (s) VTNLFE VTRE VTFE -15-10 -5 0 5 10 Population VTA VAP VTNL VT VTNLRE VTNLFE (s) VTRE VTFE (s) -5 0 5 10 15 PR VTA (l, s) VAP (l, s) VTNL VT (l, s) VTNLRE VTNLFE VTRE (l, s) VTFE -40 -20 0 20 40 2nd election VTA (s) VAP VTNL (s) VT (s) VTNLRE VTNLFE (s) VTRE VTFE (s) -10 0 10 20 30 Spending decentral. VTA VAP VTNL VT VTNLRE VTNLFE VTRE VTFE (s) -5 0 5 Sufferage VTA (s) VAP (l, s) VTNL (l, s) VT (s) VTNLRE (l, s) VTNLFE (s) VTRE VTFE (s) -30-20-10 0 1020 Time trend VTA VAP VTNL VT VTNLRE VTNLFE (s) VTRE VTFE (s) -15 -10 -5 0 5 10 Unicameralism Fig. 3 Further analyses checks, coefficient and extreme bounds plots. Note Y-axis abbreviations represent distinct model series. Horizontal lines represent extreme bounds (+/− two standard deviations). All mod-els (except those marked VTA) are limited to democracies. VTFE registered voters with fixed effects and lagged dependent variable (DV), VTRE registered voters with random effects and lagged DV, VTNLFE registered voters with fixed effects and no lagged DV, VTNLRE registered voters with random effects and no lagged DV, VT registered voter turnout, robust standard errors clustered by country, and lagged DV, VTNL registered voter turnout, robust standard errors clustered by country, and lagged DV, VAP voting aged population voter turnout, robust standard errors clustered by country, and lagged DV, VTA regis-tered voter turnout, robust standard errors clustered by country, lagged DV and including all countries. Population’s extreme bounds for VTFE and VTNLFE models excluded due to their disproportionate size (− 3666.2 and 3681.1 for VTFE) 36 We follow Kostelka (2017) in believing that the differences here require an in-depth exploration of alternative explanations of voter turnout. 57 630 Political Behavior (2023) 45:607–633 1 3 Specifically, compulsory voting increases turnout in rich states but not poor states. So too did ethnic fractionalization, economic growth, spending decentralization and inflation. Poorer states’ national voter turnout was robustly associated with boycotts (as in Western states) and concurrent elections while rich states’ turnout was not. Variation over time Finally, there may be over-time variation that is not captured by the time-trend variables in our original analyses. We therefore ran two additional series of models, one with a linear time trend and a time squared variable in all models and the second with time, time squared, and time cubed. Substantive results for the other variables were virtually identical to our main model, and the time trend variable and time cubed variables are robust in their respective model series. Take as a whole, the results of over 15 million regressions of national-level voter turnout on 70 unique variables using 16 distinct model (or sample) variations sug- gest that there are indeed a recurring series of country-level factors shaping national- level voter turnout. Some of these factors are consistent with a wide swath of the turnout literature (e.g., institutional factors like compulsory voting or concurrent elections) while others have only been included in one or two studies (e.g., socio- economic factors like economic globalization and inflation). The implications of our findings are twofold. First, our results suggest that there is room for theoretical reas- sessments of several frequently used institutional variables (e.g., electoral formulae or the number of parties). Second, our results provide a comprehensive evidence base for future turnout research. Discussion and Conclusion In this article we establish which social, institutional, and political factors driving national-level voter turnout are empirically robust using a wide vary of model speci- fications. This has three implications. First, it sheds light on what national politi- cal, institutional, or socio-economic factors are significantly associated with turn- out. Second, it provides fodder for further inductive theory development. Third, our results suggest a set of potentially useful control variables for future turnout studies. Below, we develop these three points further. First, robust results are essential to discriminate among the dozens of proposed mechanisms driving national-level voter turnout. When over a hundred factors potentially affecting turnout, it is hard to determine what robustly shapes turn- out. We collect 127 literature-derived factors and empirically analyze 70 of them. We find that some of the literature’s results are highly sensitive to small changes in model specification while others are not. Overall, we find that compulsory vot- ing, competitive elections, concurrent elections, economic globalization, inflation, previous turnout, proportional representation, spending decentralization, and some geographical dummies are robust predictors of turnout. Similar to other empirical assessments (e.g., Gassebner et al., 2013’s study of the emergence and survival of democracy) our results are humbling but provide an empirical foundation for future research. Second, from a theory-building perspective, fragile results are important because they force us to reassess both the theoretical underpinnings of our hypotheses as 58 631 1 3 Political Behavior (2023) 45:607–633 well as “reconsider whether it is theoretically reasonable to expect robustness across the various sample populations being considered” (Hafner-Burton, 2005, p. 696). Furthermore, reassessing the theories that generate our hypotheses requires taking into account other possible causes that have not received much attention in previ- ous studies. For example, several economic factors (e.g., economic globalization and inflation) are robust in our models but have yet to receive much theoretical attention in the turnout literature. We conclude by highlighting a number of possible areas for future research. For example, our data contain several different measures for the same concept. For prac- tical reasons, we have relied on the most commonly used measures. A logical next step would be to use other sources and operationalizations and compare results. Relatedly, including all or some of the remaining 57 variables included in Tables 1, 2 and 3 but not analyzed in this paper is a possibility. Furthermore, like previous studies (e.g., Hegre & Sambanis, 2006), we did not include any interaction effects in our models. Including theoretically informed interactions may provide for a more comprehensive analysis of the interplay of different explanatory factors and the way they may mediate or moderation certain relationships. Certainly, then, our analysis is limited and, by no means, definitive, but it is the largest systematic evaluation of factors shaping national-level voter turnout to date. Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s11109- 021- 09720-y. Acknowledgements We would like to thank Ignacio Lago, Ciaran O’Faircheallaigh, participants of the 2017 Australian Society of Quantitative Political Science Conference, the 2018 Doctores Miembro and former Post-Doctoral Fellows of Juan March Institute Conference, the Editor, and the anonymous peer reviewers for their valuable comments and suggestions. Replication materials are available at: https:// datav erse. harva rd. edu/ datav erse/ richa rdwfr ank. All remaining errors are our own. Author Contributions The authors are listed alphabetically and contributed equally. Funding The authors acknowledge support for this research from the Australian Research Council’s Dis- covery Project Scheme (DP150102398, DP190101978). Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Com- mons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. References Achen, C. (2001). Why lagged dependent variables can suppress the explanatory power of other inde-pendent variables. Presented at the annual meeting of the American Political Science Association. 59 632 Political Behavior (2023) 45:607–633 1 3 Bell, A., Fairbrother, M., & Jones, K. (2019). Fixed and random effects models: Making an informed choice. Quality and Quantity, 53, 1051–1074.Bell, A., & Jones, K. (2015). Explaining fixed effects: Random effects modeling of time-series cross- sectional and panel data. Political Science Research and Methods, 3(3), 133–153.Blais, A., & Carty, R. (1990). Does proportional representation foster voter turnout? European Journal of Political Research, 18(2), 167–181.Blais, A., & Dobrzynska, A. (1998). Turnout in electoral democracies. European Journal of Political Research, 33(2), 239–261.Cancela, J., & Geys, B. (2016). Explaining voter turnout: A meta-analysis of national and subnational elections. Electoral Studies, 42, 264–275.Carmignani, F., Shankar, S., Tan, E., & Tang, K. (2014). Identifying covariates of population health using extreme bound analysis. European Journal of Health Economics, 15, 515–531.Clark, T. S., & Linzer, D. A. (2015). Should I use fixed or random effects? Political Science Research and Methods, 3(2), 399–408.Colomer, J. M. (1991). Benefits and costs of voting. Electoral Studies, 10(4), 313–326. Dieleman, J. L., & Templin, T. (2014). Random-effects, fixed effects and the within-between specification for clustered data in observational health studies: A simulation study. PLoS ONE, 9(10), 1–17. Endersby, J. W., & Krieckhaus, J. T. (2008). Turnout around the globe: The influence of electoral institu-tions on national voter participation, 1972–2000. Electoral Studies, 27(4), 601–610. Fornos, C., Timothy Power, T., & Garand, J. (2004). Explaining voter turnout in Latin America, 1980 to 2000. Comparative Political Studies, 37(8), 909–940. Frank, R., & Martínez i Coma, F. (2017). How election dynamics shape perceptions of electoral integrity. Electoral Studies, 48, 143–165. Fumagalli, E., & Narciso, G. (2012). Political institutions, voter turnout, and policy outcomes. European Journal of Political Economy, 28(2), 162–173. Gassebner, M., Gutmann, J., & Voigt, S. (2016). When to expect a coup d’état? An extreme bounds anal-ysis of coup determinants. Public Choice, 169(3–4), 293–313. Gassebner, M., Lamla, M., & Vreeland, J. (2013). Extreme bounds of democracy. Journal of Conflict Resolution, 57(2), 171–197. Geys, B. (2006). Voter turnout: A review of aggregate-level research. Electoral Studies, 25(4), 637–663.Glass, G. (1976). Primary, secondary and meta-analysis of research. Educational Researcher, 5, 3–8. Hafner-Burton, E. (2005). Right or robust? The sensitive nature of repression to globalization. Journal of Peace Research, 42(6), 679–698. Hartwig, J., & Sturm, J. (2014). Robust determinants of health care expenditure growth. Applied Eco-nomics, 46(36), 4455–4474. Hegre, H., & Sambanis, N. (2006). Sensitivity analysis of empirical results on civil war onset. Journal of Conflict Resolution, 50(4), 508–535. Imai, K., & Kim, I. S. (2019). When should we use unit fixed effects regression models for causal infer-ence with longitudinal data? American Journal of Political Science, 63(2), 467–490. International IDEA. (2017). Voter turnout database. Stockholm: International Institute for Democracy and Electoral Assistance. Jackman, R. (1987). Political institutions and voter turnout in the industrialized democracies. American Political Science Review, 81(2), 405–424. Jackman, R., & Miller, R. (1995). Voter turnout in the industrial democracies during the 1980s. Com-parative Political Studies, 27(4), 467–492. Kaniovski, S., & Miller, D. C. (2006). Community size, heterogeneity, and voter turnouts. Public Choice, 129, 399–415. Kostadinova, T. (2003). Voter turnout dynamics in post-communist Europe. European Journal of Politi-cal Research, 42(6), 741–759. Kostadinova, T., & Powers, T. J. (2007). Does democratization depress participation? Voter turnout in the Latin American and Eastern European transitional democracies. Political Research Quarterly, 60(3), 363–377.Kostelka, F. (2017). Does democratic consolidation lead to a decline in voter turnout? Global evidence since 1939. American Political Science Review, 111(4), 653–667.Kuenzi, M., & Lambright, G. (2007). Voter turnout in Africa’s multiparty regimes. Comparative Political Studies, 40(6), 665–690.Laakso, M., & Taagepera, R. (1979). ‘Effective’ number of parties: A measure with application to West Europe. Comparative Political Studies, 12(1), 3–27. 60 633 1 3 Political Behavior (2023) 45:607–633 Leamer, E. (1983). Let’s take the con out of econometrics. American Economic Review, 73(1), 31–43. Leamer, E. (1985). Sensitivity analysis would help. American Economic Review, 75(3), 308–313.Levine, R., & Renelt, D. (1992). A sensitivity analysis of cross-country growth regressions. American Economic Review, 82(4), 942–963.Martínez i Coma, F. (2016). Turnout determinants in democracies and in non-democracies. Electoral Studies, 41, 50–59.Martínez i Coma, F., & Nai, A. (2017). Ethnic diversity decreases turnout. Comparative evidence from over 650 elections around the world. Electoral Studies, 49, 75–95.Martínez i Coma, F., & Trinh, M. (2017). How electoral integrity affects voter turnout in democracies. Australian Journal of Political Science, 52(1), 53–74.Martínez I Coma, F., & Morgenbesser, L. (2020). Election turnout in authoritarian regimes. Electoral Studies, 68, 102222.McDonald, M., & Popkin, S. (2001). The myth of the vanishing voter. American Political Science Review, 95(4), 963–974.McNeish, D., & Kelley, K. (2019). Fixed effects models versus mixed effects models for clustered data: Reviewing the approaches, disentangling the differences and making recommendations. Psychologi-cal Methods, 24(1), 20–35. Miller, M., Joseph, M., & Ohl, D. (2018). Are coups really contagious? An extreme bounds analysis of political diffusion. Journal of Conflict Resolution, 62(2), 410–441. Plümper, T., & Traunmüller, R. (2020). The sensitivity of sensitivity analysis. Political Science Research and Methods, 8(1), 149–159. Powell, B. (1986). American voter turnout in comparative perspective. American Political Science Review, 80(1), 17–40. Radcliff, B. (1992). The welfare state, turnout and the economy: A comparative analysis. American Polit-ical Science Review, 86(2), 444–454. Radcliff, B., & Davis, P. (2000). Labor organization and electoral participation in industrial democracies. American Journal of Political Science, 44(1), 132–141. Remmer, K. (2010). Political scale and electoral turnout: Evidence from the less industrialized world. Comparative Political Studies, 43(3), 275–303. Robbins, J., Hunter, L., & Murray, G. (2013). Voters versus terrorists: Analyzing the effect of terrorist events on voter turnout. Journal of Peace Research, 50(4), 495–508. Sala-i-Martin, X. (1997). I just ran two million regressions. American Economic Review, 87(2), 178–183.Siaroff, A., & Merer, J. (2002). Parliamentary election turnout in Europe since 1990. Political Studies, 50(5), 916–927.Smets, K., & Van Ham, C. (2013). The embarrassment of riches? A meta-analysis of individual-level research on voter turnout. Electoral Studies, 32(2), 344–359.Steiner, N. (2010). Economic globalization and voter turnout in established democracies. Electoral Stud- ies, 29(3), 444–459.Stockemer, D. (2015). District magnitude and electoral turnout: A macro-level global analysis. Acta Polit- ica, 50(1), 82–100.Stockemer, D. (2017). What affects voter turnout? A review article/meta-analysis of aggregate research. Government and Opposition, 52(4), 698–722.Stockemer, D., LaMontagne, B., & Scruggs, L. (2013). Bribes and ballots: The impact of corruption on voter turnout in democracies. International Political Science Review, 34(1), 74–90.Stockemer, D., & Khazael, S. (2014). Electoral turnout in Muslim-majority states: A macro-level panel analysis. Politics and Religion, 7, 79–99.Tavits, M. (2008). Direct presidential elections and turnout in parliamentary contests. Political Research Quarterly, 62(1), 42–54.Temple, J. (2000). Growth regressions and what the textbooks don’t tell you. Bulletin of Economic Research, 52(3), 181–205. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 61 City of Seattle 1 Seattle Districting Commission Seattle City Charter Article IV, Sec. 2, Subdivision D. COUNCIL REDISTRICTING. (1) By October 31, 2022, and of each tenth year thereafter, a five-member Districting Commission shall be appointed. The Mayor shall appoint two members, and by a two-thirds vote the City Council shall appoint two members. The fifth member shall be appointed by majority vote of the first four members. The Commission shall elect a chair from among its members. No person may serve on the Commission who is an elected official (except precinct committee officer), a registered lobbyist, a candidate for elective office, or a City employee. (2) The Districting Commission shall no later than two months after appointment, or November 30, whichever occurs first, appoint a districting master who shall be qualified by education, training and experience to draw a districting plan. If the Commission is unable to agree upon the appointment of a districting master by November 30, the Mayor shall appoint a districting master. All meetings of the Commission shall be open to the public, and the Commission shall conduct public forums around the City before adopting any plan. At least one public forum shall be held in each existing District. (3) District boundaries shall be drawn to produce compact and contiguous districts that are not gerrymandered. The population of the largest district shall exceed the population of the smallest by no more than one percent. To the extent practical, district boundaries shall follow existing District boundaries, recognized waterways and geographic boundaries, and Seattle communities and neighborhoods. In drawing the plan, neither the Commission nor the districting master shall consider the residence of any person. (4) The districting master shall draw a districting plan for the City, and submit it to the districting Commission by March 15 of the year following their appointment. The Commission shall develop, approve by majority vote, and make public a draft districts proposal and then after public comment, approve by majority vote a final districts plan; and shall have all powers reasonably necessary to carry out its purpose, may employ experts, consultants and attorneys not employed by the city, and shall prepare financial statements and compose and turn over to the City Clerk an official record of all relevant information used. Upon adoption, the districting plan shall be filed with the City Clerk. The plan shall become effective upon filing and cannot be amended by the City Council except to correct data errors upon request by the districting Commission. (5) The City Council shall take all steps necessary to enable the Commission to function, including appropriations sufficient to provide a reasonable per diem for Commission members, compensation for staff and contractors, and reasonable expenses. Current as of: September 2018 62 City of Seattle 2 www.localredistricting.org 63 SAN DIEGO CITY CHARTER II. 1 ARTICLE II NOMINATIONS AND ELECTIONS SECTION 4: DISTRICTS ESTABLISHED For the purpose of electing members of the Council, the City shall be divided into nine Council districts as nearly equal in population as practicable. The boundaries of such districts shall be subject to alteration and change under the provisions of this Charter. In any redistricting plan adopted by the Redistricting Commission pursuant to Section 5.1 or ordinance adopted by the Council establishing, changing or altering the boundaries of any Council district, the redistricting plan or ordinance may describe the new boundaries by reference to a map on file in the office of the City Clerk; a metes and bounds description of the new boundaries need not be contained in the redistricting plan or ordinance. (Amendment voted 03-10-1953; effective 04-20-1953.) (Amendment voted 09-17-1963; effective 02-11-1964.) (Amendment voted 11-06-1990; effective 02-19-1991.) (Amendment voted 06-02-1992; effective 07- 13-1992.) (Amendment voted 06-08-2010; effective 07- 30-2010.) (Amendment voted 06-07-2016; effective 07- 18-2016.) SECTION 5: REDISTRICTING In the event that any voting precinct established is located partly within two or more Council districts, the precinct shall be allocated to the Council district in which a majority of the voters within the precinct resides, and the district boundaries shall be changed accordingly. The City shall be redistricted pursuant to Section 5.1 of this Charter at least once in every ten years, but no later than nine months following the City’s receipt of the final Federal Decennial Census information. The term “Federal Decennial Census,” as used in this Charter, shall mean the national decennial census taken under the direction of the United States Congress at the beginning of each decade. Any territory hereafter annexed to or consolidated with The City of San Diego shall at the time of such annexation or consolidation be added to an adjacent district or districts by an ordinance of the Council. However, if any territory annexed, deannexed or consolidated upsets the approximate equality of the populations of the established districts, a redistricting shall be conducted pursuant to Section 5.1 of this Charter, except that the nomination period for appointment to the Redistricting Commission shall commence on the May 1 immediately succeeding the annexation, deannexation or consolidation and the Redistricting Commission shall be constituted no later than the next November 1. In any redistricting, the districts shall be comprised of contiguous territory and made as equal in population as shown by 64 SAN DIEGO CITY CHARTER II. 2 the census reports, and as geographically compact as possible, and the districts formed shall, as far as possible, be bounded by natural boundaries, by street lines and/or by City boundary lines. (Amendment voted 06-02-1992; effective 07- 13-1992.) (Amendment voted 06-07-2016; effective 07- 18-2016.) SECTION 5.1: REDISTRICTING COMMISSION The members of the City Council shall be elected by districts. Subject to the provisions of the City Charter relating to referendum and initiative powers of the people, the sole and exclusive authority to adopt plans which specify the boundaries of districts for the City Council is vested in the Redistricting Commission, to be established by this Section. Commencing in the year following the year in which the Federal Decennial Census is taken at the beginning of each decade, the Redistricting Commission shall adopt plans that redistrict the City into nine Council districts designated by numbers 1 to 9, inclusive. Those districts shall be used for all elections of Council members, including their recall, and for filling any vacancy in the office of member of the Council (and until new districts are established). No change in the boundary or location of any district by redistricting as herein provided shall operate to abolish or terminate the term of office of any member of the Council prior to the expiration of the term of office for which such member was elected. Districts formed by the Redistricting Commission shall each contain, as nearly as practicable, one-ninth of the total population of the City as shown by the Federal Decennial Census immediately preceding such formation of districts. Each redistricting plan shall provide fair and effective representation for all citizens of the City, including racial, ethnic, and language minorities, and be in conformance with the requirements of the United States Constitution and federal statutes. To the extent it is practical to do so, districts shall: preserve identifiable communities of interest; be geographically compact - populous contiguous territory shall not be bypassed to reach distant populous areas; be composed of whole census units as developed by the United States Bureau of the Census; be composed of contiguous territory with reasonable access between population centers in the district; and not be drawn for the purpose of advantaging or protecting incumbents. The Redistricting Commission shall be composed of nine members who shall be appointed by a panel of three retired judges who served in any of the following courts: the Superior Court of the State of California, an appellate court of the State of California, or a U.S. District Court located within California. Names of the retired judges willing to serve will be submitted to the City Clerk and drawn at random by the City Clerk, using procedures for judicial nominees and 65 SAN DIEGO CITY CHARTER II. 3 appointees as set forth in the San Diego Municipal Code. The City Clerk shall also draw at random the name of one additional retired judge to be designated as an alternate, who will be appointed to serve on the panel if another member is unable or unwilling to serve. The term “Appointing Authority,” as used herein below, shall refer to the panel of retired judges acting to appoint the Redistricting Commission pursuant to the provisions of this paragraph. The City Clerk shall solicit nominations for appointment to the Redistricting Commission in accordance with this Section and shall distribute to the news media the announcement of a sixty-day nomination period (which shall commence on May 1 of every year in which a Federal Decennial Census is taken) and the guidelines for selection of Commission members. Individuals or organizations desiring to nominate persons for appointment to the Commission shall submit application materials to the City Clerk within the nominating period, using procedures set forth in the San Diego Municipal Code. The City Clerk shall transmit the names and information regarding all nominees with the names of nominating individuals and organizations to the Appointing Authority immediately upon the close of nominations. After receiving Commission member applications from the City Clerk, the three members of the Appointing Authority shall hold a public meeting to appoint the nine Commission members and two alternates. The public meeting shall be held as promptly as possible to ensure the appointments are timely made. The Appointing Authority shall appoint the members constituting the Commission no later than November 1 of every year in which a Federal Decennial Census is taken. In the event that a complete panel of three retired judges is unable or unwilling to serve as the Appointing Authority, the City Clerk will serve as the Appointing Authority. The City Clerk will conduct a ministerial review of Commission member applications to determine which persons are qualified to serve, using the requirements of this Charter and procedures set forth in the San Diego Municipal Code. After all qualified applicants are identified and notified, the City Clerk will randomly select the Commission members and alternates from the pool of qualified applications in a public place. The Appointing Authority shall appoint members who will give the Redistricting Commission geographic, social and ethnic diversity, and who, in the Appointing Authority’s judgement, have a high degree of competency to carry out the responsibilities of the Commission. The appointees shall include individuals with a demonstrated capacity to serve with impartiality in a nonpartisan role. The Appointing Authority shall attempt to appoint one Commission member from each of the nine Council districts to the extent practicable, given the other requirements of this Charter Section, and considering the extent of the applicant pool and an individual’s qualifications to 66 SAN DIEGO CITY CHARTER II. 4 serve. The Appointing Authority shall also attempt to appoint Commission members who possess working knowledge of the geography and neighborhoods of The City of San Diego. Each member of the Commission shall be registered to vote in The City of San Diego. The Appointing Authority shall also appoint two alternate Commission members from the same applicant pool, who will be available to serve if a Commission member cannot serve or resigns. The alternates shall have a duty to remain informed of Commission business so they are prepared to serve if called upon to do so. Persons who accept appointment as members of the Commission, at the time of their appointment, shall file a written declaration with the City Clerk stating that within five years of the Commission’s adoption of a final redistricting plan, they will not seek election to a San Diego City public office. Alternates shall sign the declaration if and when they become members of the Commission. The members of the Redistricting Commission, and the alternates, shall serve until the redistricting plan adopted by the Commission becomes effective and the referendary deadline for the Final Redistricting Plan has passed. If the Final Redistricting Plan is rejected by referendum or by a legal challenge, members of the Commission shall resume their service and shall create a new plan pursuant to the criteria set forth in Sections 5 and 5.1. Within thirty days after the Commission members are appointed, the Commission shall hold its first meeting at a time and place designated by the City Clerk. Within sixty days after the Commission members are appointed, the Commission shall adopt a budget and submit it to the Appointing Authority. If approved, the budget shall be forwarded to the City Council for its prompt consideration. The City Council shall appropriate funds to the Commission and to the City Clerk adequate to carry out their duties under this Section. All Commission meetings shall be open to the public and Commission records, data and plans shall be available, at no charge, for public inspection during normal business hours in the office of the City Clerk. Copies of records and plans shall be provided, for a reasonable fee, to any interested person. The Commission shall elect a chair and a vice chair and shall employ a chief of staff, who shall serve at the Commission’s pleasure, exempt from Civil Service, and shall contract for needed staff, technical consultants and services, using existing City staff to the extent possible. Aye votes by six members of the Commission shall be required for the appointment of its chief of staff, the election of its chair, and the adoption of the Final Redistricting Plan. A majority vote of the Commission shall be required for all other actions. A majority of the entire Commission shall constitute a quorum for the transaction of business or exercise of any power of the Commission. 67 SAN DIEGO CITY CHARTER II. 5 The Commission shall make every reasonable effort to afford maximum public access to its proceedings. It shall solicit public comment and shall hold at least nine public hearings in various geographic areas of the City before the preparation of a preliminary redistricting plan. At least thirty days prior to the adoption of a final plan, the Commission shall file a preliminary plan with the City Clerk, along with a written statement of findings and reasons for adoption, which shall include the criteria employed in the process and a full analysis and explanation of decisions made by the Commission. During the thirty day period after such filing, the Commission shall hold at least five public hearings in various geographic areas of the City before it adopts a final plan. The Final Redistricting Plan shall be effective thirty days after adoption by the Commission and shall be subject to the right of referendum in the same manner as are ordinances of the City Council. If rejected by referendum, the same Commission shall be empaneled to create a new plan pursuant to the criteria set forth in Sections 5 and 5.1. The Final Redistricting Plan document, including all maps, will be final as set forth herein; however, the boundaries of the Council districts in the Final Redistricting Plan shall not be adjusted and effective until after the next regularly scheduled general election for Council seats following the redistricting. If any part of these amendments to Sections 4, 5, or 5.1 of the Charter or their application to any person or circumstances is held invalid, the invalidity shall not affect other provisions or applications which reasonably can be given effect without the invalid provision or application. (Addition voted 06-02-1992; effective 07-13- 1992.) (Amendment voted 06-08-2010; effective 07- 30-2010.) (Amendment voted 06-07-2016; effective 07- 18-2016.) SECTION 6: QUALIFIED ELECTORS The qualifications of an elector at any election held in the City under the provisions of this Charter shall be the same as those prescribed by the general law of the State for the qualification of electors at General State Elections. No person shall be eligible to vote at such City election until he has conformed to the general State law governing the registration of voters. SECTION 7: ELECTIVE OFFICERS RESIDENCY REQUIREMENT An elective officer of the City shall be a resident and elector of the City. In addition, every Council member shall be an actual resident and elector of the district from which the Council member is nominated. The office of a Councilmember shall be vacated if he or she moves from the district from which the Councilmember was elected. Redistricting that occurs during a Councilmember’s term shall not operate to create a vacancy. The Council shall establish by ordinance 68 SAN DIEGO CITY CHARTER II. 6 minimum length of residency requirements for candidacy to elective office, whether by appointment or election. (Amendment voted 09-17-1963; effective 02-11-1964.) (Amendment voted 11-04-1969; effective 01-29-1970.) (Amendment voted 11-06-1979; effective 12-17-1979.) (Amendment voted 11-08-2016; effective 12-19-2016.) SECTION 8: ELECTION CODE Within ninety (90) days after this amendment has been ratified by the State Legislature the Council shall adopt an election code ordinance, providing an adequate and complete procedure to govern municipal elections, including the nomination of candidates for all elective offices. All elections provided for by this charter, whether for choice of officers or submission of questions to the voters, shall be conducted in the manner prescribed by said election code ordinance. (Amendment voted 04-22-1941; effective 05-08-1941.) SECTION 9: NOMINATIONS Nominations of candidates for all elective offices shall be made in the manner prescribed by the election code ordinance provided for in Section 8 of this article. (Amendment voted 04-22-1941; effective 05-08-1941.) SECTION 10: ELECTIONS Elective officers of the City shall be nominated and elected by all of the electors of the City except that City Council members shall be nominated and elected by the electors of the district for which elective office they are a candidate. Commencing with the year 1996, the municipal primary elections to the office of Council member for Districts 1, 3, 5, and 7 shall be held on same date in each election year as the California State primary election, and the general municipal election for these offices shall be held on the same date as the California State general election for that year. Commencing with the year 2012, the election to the office of Council member for District 9 shall be held on the same date as the election to the office of Council member for Districts 1, 3, 5, and 7. Commencing with the year 1998, the municipal primary elections to the offices of Council member for Districts 2, 4, 6, and 8 shall be held on same date in each election year as the California State primary election, and the general municipal election for these offices shall be held on the same date as the California State general election for that year. Commencing with the next municipal primary and general elections following the redistricting occurring after the 2010 national decennial census, and every four years thereafter, the municipal primary and general elections to the office of Council District 9 shall be held. Commencing with the year 1984 the elections to the offices of Mayor and City 69 SAN DIEGO CITY CHARTER II.7 Attorney shall be held every four (4) years. The municipal primary election for the offices of Mayor and City Attorney shall be held on the same date in each election year as the California State primary election, and the general municipal election for these offices shall be held on the same date as the California State general election for that year. All other municipal elections which may be held under this Charter shall be known as special municipal elections. All elective officers of the City shall be nominated at the municipal primary election. The two candidates receiving the highest number of votes for a particular elective office at the primary shall be the candidates, and only candidates, for such office and the names of only those two candidates shall be printed upon the ballots to be used at the general municipal election. In the event only one candidate has qualified for the ballot in the municipal primary election for a particular elective office, the sole qualified candidate receiving votes in the municipal primary election shall be deemed to be, and declared by the Council to be, elected to such office after the primary election results are certified. At the general municipal election held for the purpose of electing Council members, the electors of each Council district shall select from among the candidates chosen at the primary election in that district one candidate for the office of the Council member whose term expires the succeeding December. At the general municipal election held for the purpose of electing any other elective officer, there shall be chosen by all of the electors of the whole City from among the candidates chosen at the primary one candidate to succeed any other elective officer whose term expires in December succeeding the election. After the result of an election for any office is declared, or when an appointment is made, the City Clerk, under his or her hand and official seal, shall issue a certificate therefor, and shall deliver the same immediately to the person elected or appointed, and such person must within ten days after receiving such certificate file his official bond, if one be required for his office, and take and subscribe to the oath of office required of him by this Charter, which oath must be filed with the City Clerk. (Amendment voted 04-22-1941; effective 05-08-1941.) (Amendment voted 06-05-1956; effective 01-10-1957.) (Amendment voted 11-06-1962; effective 01-21-1963.) (Amendment voted 11-04-1975; effective 12-01-1975.) (Amendment voted 11-08-1988; effective 04-03-1989.) (Amendment voted 11-03-1992; effective 12- 18-1992.) (Amendment voted 06-08-2010; effective 07- 30-2010.) (Amendment voted 06-03-2014; effective 07- 24-2014.) (Amendment voted 11-08-2016; effective 12- 19-2016.) 70 Memorandum REPORT TO:Study Commission SUBJECT:Discussion with Jeff Krauss and Danielle Rogers on City Boards MEETING DATE:April 2, 2026 AGENDA ITEM TYPE:Citizen Advisory Board/Commission RECOMMENDATION:Discuss with Former Mayor and City Commissioner Jeff Krauss and the Economic Vitality Board Chair Danielle Rogers on the effectiveness and change of Bozeman City Advisory Boards. 71 Memorandum REPORT TO:Study Commission FROM:Caeleb Heinen, Recording Secretary Mike Maas, Ex Officio Dan Clark, MSU Local Government Center SUBJECT:Learning session for Neighborhood Associations and Boards MEETING DATE: April 2, 2026 AGENDA ITEM TYPE: Citizen Advisory Board/Commission RECOMMENDATION: Participate in a Learning Session with Montana State University Local Government Center's Dan Clark on the specificities of Neighborhood Associations and Boards. STRATEGIC PLAN: 1.1 Outreach: Continue to strengthen and innovate in how we deliver information to the community and our partners. BACKGROUND: Reference Materials are attached. For Neighborhood Associations: Recommended INC Charter Language For Wards/Districts on Voter Turnout and Gerrymandering: Comparison City Chart (with Voter Turnout) Voter Turnout in At Large vs. Wards Correlates of Voter Turnout - Political Behavior Seattle Districting Commission San Diego Charter Redistricting UNRESOLVED ISSUES: None Identified ALTERNATIVES: As per the Study Commission FISCAL EFFECTS: None Identified Attachments: Recommended INC Charter Language.pdf Report compiled on: March 25, 2026 72 73 Final INC recommended Charter Language Approved 12/11/25 (a) Intent: The citizens of Bozeman, and the City of Bozeman, value the contribution neighborhoods can make to the governance of the city. Therefore, it is the purpose of this article to strengthen neighborhood participation where it exists, and to encourage and support neighborhood participation where it does not yet exist. (b) Purpose: There shall be a City program that supports the creation and maintenance of resident-led neighborhood associations. The associations shall act in an advisory capacity, and may contribute information, opinions, advice, suggestions and recommendations to the City Commission, City Manager and other City advisory bodies on all governmental affairs and services having an effect on the area the neighborhood association represents, including, but not limited to, public finance, public works, public safety, planning and zoning, and public health and sanitation. (c) Recognition of neighborhood associations. The Inter-Neighborhood Council shall establish minimum recognition standards for neighborhood associations, to be adopted by the City Commission by ordinance. These standards shall include, but not be limited to: (1) clear geographic boundaries; (2) adherence to established minimum by-laws that ensure voting capabilities by a neighborhood association’s selected Inter-Neighborhood Council representative; (3) periodic meetings, including an annual meeting; (4) copy of the by-laws and all amendments filed with the City; (5) the definition of membership eligibility includes all residents, property owners, business owners, and nonprofit organizations located within a given neighborhood association’s boundaries; and (6) demonstrated means of communication with all members in a neighborhood association. (d) Minimum Standards. A neighborhood association must meet and continue to maintain conformity with the minimum standards as established by ordinance in order to be recognized by the City of Bozeman and to be eligible to select members and representatives to the Inter-Neighborhood Council. Neighborhood associations existing on the date of the enactment of this charter shall have one year after the enactment of said City ordinance to come into compliance. 74 (e) Inter-Neighborhood Council. There is hereby established, an Inter-Neighborhood Council (INC) to be composed of representatives selected by each formally recognized neighborhood association. (1) The INC shall provide a public forum for formally recognized Neighborhood Association representatives to meet, share information, and make recommendations to the City Commission and City Manager. Recommendations made by the INC may be related to neighborhood-level and/or city-wide issues in order to advocate for their resident’s concerns. INC recommendations do not preclude a neighborhood association from taking its concerns directly to the Mayor/City Commission and/or City Manager. (2) The INC shall meet on a regular basis to understand and address city-wide concerns, examine current or proposed city policy, and foster dialogue between neighborhoods and with City leadership. (3) The INC shall adopt by-laws governing the conduct of their business. Such by-laws shall be adopted by the City Commission, by ordinance. A vacancy on the INC shall be filled only by the affected neighborhood association. The INC may include a City Commissioner as a non-voting member of the INC. (f) Neighborhoods Coordinator. The City shall retain a full-time paid staff member to coordinate with neighborhood associations and the Inter-Neighborhood Council. This staff member shall be under the City Manager’s Department and work solely on the neighborhoods program. 75 Memorandum REPORT TO:Study Commission FROM:Caeleb Heinen, Recording Secretary Mike Maas, Ex Officio SUBJECT:Future Education Discussion and Needs for April 15 Meeting MEETING DATE:April 2, 2026 AGENDA ITEM TYPE:Citizen Advisory Board/Commission RECOMMENDATION:Education process and needs for April 15 for the following topics around community representation and involvement: wards/districts neighborhood associations advisory boards STRATEGIC PLAN:1.1 Outreach: Continue to strengthen and innovate in how we deliver information to the community and our partners. BACKGROUND:This is an opportunity for Study Commissioners to discuss what further materials or education they will need on the aforementioned topic. UNRESOLVED ISSUES:None identified ALTERNATIVES:As per the Study Commission FISCAL EFFECTS:None identified Report compiled on: March 25, 2026 76