Television ratings during tornado warnings in eastern Iowa

Every spring it’s inevitable meteorologists will cut into regularly scheduled programming to provide life-saving information when tornadoes threaten communities. Whether it’s a big sports game, a hit reality TV show or the season finale of a drama, in most television markets tornado warnings will supersede that content.

A small sample of the social media feedback during tornado warning coverage.

Viewers tend to show their disdain for the fact meteorologists are on air, and it’s also coming in over the phone and via email. Many people saying things like “get off the TV” and “no one cares.” So that begs the question, is that accurate? Do people wish we were off the air, or are people tuning into severe weather coverage?

RESEARCH METHODOLOGY

To answer the question of whether or not people are tuning in, the data will be television ratings. For this project, we will be taking COMSCORE viewership data during a tornado event and will compare it to the week prior, at the same time.

For this research, the focus will be weekday evening events affecting Johnson County to hone in on a specific geographic area which may help control some of the data.

RESEARCH LIMITATIONS

While ratings data are great source of information, it doesn’t come without its limitation. Picking specific days with tornado warnings, it can be difficult to work around season finales or important episodes which may draw more viewership and skew data. Additionally, if there is an show that has historically low viewership, when comparing it to tornado coverage the week later, that might also skew the data in the opposite way.

Also, active weather in another part of the market, say Linn County or Black Hawk County, that may drive more viewership.

Finally, the way COMSCORE data are collected is not necessarily real time, but in 15-minute chunks. This may affect short-duration events or may not reflect significant moments that may last just a few minutes.

RESEARCH DATES

For this project, tornado warnings over the last two years are chosen. This includes May 2, 2018; Aug. 28, 2018; October 9, 2018; May 24, 2019; May 29, 2019; and Sept. 12, 2019.

All of these cases had an affect on Johnson County, with the exception of Sept.12 which will be used as a control.

In all six of the cases outlined, coverage at KGAN-TV featured live storm chaser video inside of tornado warning polygons, viewer photos coming in and meteorologists in front of the radar. In all but one case (May 2, 2018), there were confirmed tornadoes in the area. The May 24, 2019 case featured live tornado video for more than 10 minutes approaching a major population center.

May 2, 2018, Aug. 28, 2018, May 24, 2019 and May 29, 2019 featured tornado warnings for the Iowa City metropolitan area, the largest city in Johnson County.

DATA ANALYSIS

In every case, viewership during tornado warnings saw big gains over the same period a week prior.

For these examples, viewership across the whole time of tornado warning coverage is compared to the entire viewership during the same period the week before. This is an average of the entire event, which may be affected by different shows ending and starting when compared to the week with tornadoes.

At KGAN-TV, in every single case, viewership saw gains during the week of tornado coverage compared to the week prior - some by significant amounts. On May 29. 2019, the three-hour marathon coverage drew a 45 percent gain in viewership. Part of that could be the fact seven counties were touched by tornado warnings in the viewing area.

May 24, 2019 was the least gain overall, however this was competing against the CBS “Royal” Special on the British Royal Family, "Meghan and Harry Plus One,” which had high ratings nationally.

Viewership increased 58 percent while a tornado was on television approaching Iowa City.

When the May 24, 2019 event is broken down into the 15-minute blocks, that is where you see the significant gains compared to the week prior. When starting the 6 p.m. news prior to a tornado warning being issued, viewership was a bit less than the week before, also competing against the 6 p.m. news.

As storms were forming and Meteorologist Nick Stewart was showing an impressive wall cloud between 6:15-6:30 p.m., viewership was 8 percent higher. Between 6:30-6:45 p.m. while a tornado warning was issued and KGAN-TV was airing storm chaser video of the storm, viewership was 18 percent higher.

The big gains came between 6:45-7:00 p.m. as a very photogenic tornado was on air, with a meteorologist chasing it explaining street-by-street where it was. Viewership was up 58 percent over Entertainment Tonight the week prior.

Viewership during tornado warning coverage on May 29 was more than double the week before not only on KGAN, but across the entire market.

May 29, 2019 was a rather special event in this study. There were nearly three hours of coverage on this date with warning after warning issued. There were also several confirmed tornadoes. Of particular note was the fact at least a part of Johnson County was under a tornado warning from 7:11-9:03 p.m.

When breaking down the 15-minute blocks, there was a significant gain across the board in viewership. Between 8-8:45 p.m. viewership doubled not only on KGAN, but in the entire market.

On KGAN (CBS), the viewership was double a fan-favorite “The Amazing Race.” Viewership remained strong during the entire event and had even seen gains as more warnings were issued.

SUMMARY

The KGAN Road Warrior storm chaser in front of a severe thunderstorm.

In summary, every time tornado warning coverage preempted regularly scheduled programming there were gains in viewership, at times significant. The mean for KGAN was a 29 percent gain when looking at the total event compared to the same time.

In a few cases viewership actually doubled.

FUTURE WORK

Looking ahead, more cases will be added to this research including events in 2020, as well as events in 2016 and 2017 while data was fairly high resolution. Additionally, data from the week after a tornado event would also be thrown in to compare to help get a better idea of the change in viewership.

A potential expansion of this project would also be factoring and comparing tornado days with live storm chasing video and events without live video to see if that has any effect in drawing in viewers.