Determinants of consumption for regional sports network programming: an examination of inheritance effects, lead-in, lead-out, and game viewership

ABSTRACT

HIGHLIGHTS

• Estimate viewership determinants for sequenced live sports programming broadcast on Regional Sports Networks.

• Statistically significant evidence of audience carryover from each broadcast to the next.

• Viewership preferences vary substantially by program type, suggesting unique consumption preferences.

• Local team performance impacts viewership for the subsequent pre-game show broadcast.

The financial health of major North American professional sports franchises is closely tied to Regional Sports Networks. These networks purchase local market broadcast rights from franchises and then create sequenced pre-game and post-game show content bookending each live contest. Despite the financial importance of Regional Sports Networks, no study has empirically examined the drivers of viewership for this sequential content. We create a sample of sequenced pre-game show, game broadcast, and post-game show ratings and uncover statistically significant audience carryover from each broadcast to the next. We also demonstrate substantial variance in the viewership determinants for each live program, suggesting viewers have different consumption preferences for each program type. We further illustrate that local team performance relative to expectations significantly impacts ratings for the subsequent non-adjacent pre-game show broadcast. The results are crucial to broadcast rights valuation, advertising values, and programming strategy.

KEYWORDS:

• Regional sports networks
• television viewership
• National Basketball Association
• audience carryover
• inheritance effects

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The National Football League sells all broadcast rights nationally and does not allow individual franchises the ability to sell local broadcast rights.

² We test for potential endogeneity in audience carryover using a Wu-Hausman test, but do not find statistical evidence supporting endogeneity (p = 0.6350). This indicates OLS is consistent and a correction is not needed.

³ We again conduct a Wu-Hausman test and fail to find statistical evidence of endogeneity (p = 0.3299).

⁴ The correlation between ClosingLine and SpursElo is −0.2637. The correlation between ClosingLine and OppoentElo is 0.6387. Both values are within accepted limits regarding collinearity (Kutner et al., 2004).

⁵ While there is a natural positive correlation between the number of superstar players a team employs and team quality, the correlation between AllNBA and SpursElo (0.1624), OpponentElo (0.4169), and CombinedElo (0.4662) are within acceptable limits regarding collinearity (Kutner et al., 2004).

⁶ To ensure the effect of AllNBA is not being driven by its correlation with the absolute quality measures, we estimate alternative models (results not presented due to brevity) where the absolute quality measures are removed. The effect of AllNBA remains negative and fails to reach statistical significance. This suggests pre-game show viewership is driven by contest quality and opponent team quality (depending on the specification) and not star players.