Measuring Political Airtime in Television News: Protocol, the Denominator Problem, and Reliability Design

A coding protocol for measuring the allocation of speaking time to political actors, and for identifying patterns of editorial framing, in the central news programmes of national-coverage television in Serbia. The protocol follows the OSCE/ODIHR standard in separating measured time from interpretive assessment, and is designed so that every published figure can be traced to a second of recorded broadcast. The note reports one methodological finding that applies to airtime monitoring generally, whether automated or manual: the choice of denominator changes the headline figure by a factor of nearly five, and the denominator in use is rarely declared. In one fully adjudicated broadcast, the share of airtime held by the governing side is 100 %, 66.7 %, 39.7 % or 20.2 % depending on the denominator applied. All four are arithmetically correct. The note also states, without qualification, that an inter-coder agreement coefficient has not been computed, and sets out which parts of the measurement rest on human judgement and which do not. Deposited with the note is the sampling design for the campaign preceding the Serbian parliamentary elections of 25 October 2026, fixed in advance of the first broadcast it covers. The sampling plan is written in Serbian; this record is described in English.