For the first time in history, more than half of U.S. households (50.8%) have cut the cord to their landline telephones and rely exclusively on cellphones.
This continuing shift to cell-phone-only (CPO) households has major ramifications for Nielsen and other researchers that use random-digit dialing to landlines to recruit survey participants, especially among younger demos less likely to have a landline.
The latest data from the Centers for Disease Control and Prevention’s National Health Interview Survey, conducted in the second half of 2016, found a 2.5% year-over-year increase in homes that didn’t have a landline and where at least one resident had a mobile phone. More than 123 million adults live in households with at least one wireless phone, the study shows, while around 39% of homes used both landlines and cellphones. Another 3.2% of households had no phone service at all.
The CDC’s in-person survey included 19,956 households and is considered the gold-standard benchmark for tracking the number of homes without a landline.
The data shows much higher incidences of CPO households among younger demos and those who rent their homes. For instance, more than 70% of adults 25-34 live in households with wireless-only service and 71.5% of renters are cellphone only.
Like Arbitron before it, Nielsen has been ramping up the number of participants it recruits who live in households that only have cellphones. That’s important because persons living in CPO households behave differently than persons living in landline households. And some landline households never answer their landline phone and only use their cells.
Because federal law prohibits the use of auto-dialing equipment to call cellphones, reaching CPO households requires more expensive address-based sampling.
Back in Dec. 2011, Arbitron said goodbye to a hybrid system of random-digit dialing to landlines and address-based sampling and switched to a new methodology that invited households into its PPM panels based solely on their address. By the end of 2013, after Nielsen acquired Arbitron, PPM panels were fully selected from an address-based frame. The upside is that cellphone households are fully represented in the frame without needing to be separately sampled or targeted. By 2014, roughly 45% of addresses in Nielsen’s PPM sample frame couldn’t be matched to a landline phone number, a number that jumped to 58% this year. The number is consistent with the CDC data since what Nielsen calls “unmatched” addresses are mainly CPO households but also include homes with no phones and households that have a landline number but can’t be matched to that number through available databases.
As it adds more “unmatched” households to its sample, Nielsen has had to increase the amount of sample it has recruited in-person. In PPM markets, the company says the amount of panelists recruited by knocking on doors has jumped from 10% to 25%. The combination of an address-based sample and in-person recruiting is intended to produce a PPM sample that better represents the demographics of the market being measured, also known as proportionality.
Meanwhile in diary markets, a different methodology is used to include the growing portion of the population that can’t be reached via a landline phone. Arbitron switched to a hybrid sample frame in 2009-10 that uses an address-based sample frame which includes CPO households to supplement the random-digit-dialing landline frame. That started in 2010 with just 9% of the diary sample coming from CPO households, using addresses as the sampling starting point. But it nearly doubled to 17% later that year. Since then, as more consumers have cut the landline cord, Nielsen increased the percentage of address-based sample with two big rises, in fall 2016 and spring 2017.
Nielsen says it expects an average 68% of its diary sample to be recruited from an address-based frame across all of its diary markets in the spring 2017 diary survey. That, it says, adds up to a 70% increase in the amount of CPO sample since 2015. But the percentage in individual markets varies based on historical market performance and regional differences in phone status from CDC data.
Nielsen says the increases have “helped us better represent younger persons who are more likely to live in cellphone households,” and that the fall 2016 increase helped drive a 9-point increase in persons 18-34 proportionality compared to fall 2015.