Do Bangladeshi Governing Parties Keep Their Election Pledges? A Registered Report and Coding Protocol for a 1991–2026 Pledge-Fulfillment Audit

Public discussion of Bangladeshi politics often asserts that governing parties abandon their election promises, but the evidence I could locate is sectoral or journalistic rather than a systematic, pledge-level, multi-election measurement. This report specifies, before any data are coded, an audit of the fulfillment of verifiable manifesto pledges by governing parties following the 1991, June 1996, 2001, 2008, 2014 and 2018 elections, with the 2024 and 2026 elections treated as interim cases. Each pledge is classified by policy domain and specificity and scored as fully fulfilled, partially fulfilled, not fulfilled, or not assessable, using dated documentary evidence and two independent coders blinded to party where feasible. Four hypotheses with explicit falsification rules are stated:H1: overall fulfillment is not substantially below the cross-national benchmark reported by Thomson et al. (2017).H2: fulfillment is selective, higher for economic and infrastructure pledges than for governance and accountability pledges.H3: fulfillment differs between caretaker-era and later governments.H4: pledges of governing parties are fulfilled more often than pledges of non-governing parties. The main threats to validity are the small number of governing terms (at most six), unequal evidence availability across policy domains, and access to pre-2008 manifesto texts, which I could not locate online. The contribution at this stage is the design, codebook, decision rules and open analysis code; findings will follow data collection. Files: the manuscript (PDF), the analysis script (pledge_analysis.py) and the empty coding template (pledges_coded_template.csv).