The hearing majority of captioning viewers

SUMMARY – Hearing people are now almost certainly the main audience of closed-captioning, outnumbering deaf viewers by a factor of several hundred.

Introduction

To watch captioned TV in North America, you need a decoder to make closed-captions visible. Until 1993, virtually the only way to watch television captions was to buy a separate set-top decoder.

The largest consistent estimate I have read of decoder sales in all of North America in the entirety of the first 13 years of captioning is 300,000. (I don’t have a citation.) Statistics Canada (in “Selected Characteristics of Persons with Disabilities Residing in Households,” 1994) states that there were 15,575 decoder users in Canada (and 10,240 who needed decoders but didn’t have them, a number not considered here).

Further statistics on the number of users of set-top decoders in the U.S. just after the decoder law kicked in are now available courtesy of Stephen Kaye of the Disability Statistics Center, University of California San Francisco. 15% of deaf people and 1.4% of hard-of-hearing people report using caption decoders. That amounts to 48,000 deaf people plus 127,000 hard-of-hearing, or 175,000 total. (“This is based on 1994–1995 data from the National Health Interview Survey on Disability,” Kaye says.)

Summing up the numbers, then:

300,000 set-top decoders in North America
284,425, or 94.8%, are in the U.S.
15,575, or 5.2%, are in Canada by StatsCan 1994 estimate

Let’s assume that each of those old set-top decoders is watched by one person – a conservative estimate, since many households consist of more than one person. Let’s also assume that all those people are deaf or hard-of-hearing – also a conservative assumption, since NCI propaganda stated that the largest single group buying decoders in the late 1980s and early 1990s consisted of people learning English as a second language, nearly all of whom are hearing, and since a small number of hearing people have been watching captioning since its early days.

Television sales

The Electronics Industries Association estimates the following sales of TV sets:

23,005,000 in 1993 (we’ll be using half this number: The decoder law was not in effect for the first half of 1993, as you will see below)
An estimated 24,820,000 in 1994
An estimated 25,600,000 in 1995

The EIA started charging a fee for its figures in 1996, so I extrapolate from 1995 data and assume a conservative 25,000,000 sales every year from 1996 onward. Assume 90% of those are 13 inches in size or larger.

Decoder requirements

The Television Decoder Circuitry Act came into effect in mid-1993 in the U.S. It required new TV sets with screens 13 inches or larger to contain caption-decoding chips. So half the TVs sold in 1993 were subject to that law and all TVs made thereafter (except for those smaller than 13 inches in size). The law also applies to computer video boards and other devices capable of displaying TV signals, but those devices are not counted here. A few very small TV sets, like some Zeniths, nonetheless come with decoder chips.

There is no requirement in Canada, but since split runs of Canadian-only sets are rare (apparently only Samsung manufactures such split runs, and even then not always), in practice Canadian sets show the same distribution of decoder equipment.

U.S. calculations

Looking at the numbers:

In the U.S., a total of 198,425,000 TV sets were sold from mid-1993 to 2000.
Assume 90% of those sets are 13 inches in size or larger. That means 168,230,250 decoder-equipped TV sets were sold since the decoder law kicked in, which is 583 times the number of set-top decoders sold in the U.S.
If 90% of those sets are watched by hearing people (a conservative estimate), then 151,407,225 hearing people have decoder-equipped TVs, which is 532 times the number of set-top decoders.
If Kaye’s figures on decoder usage are correct (175,000 users in 1994–1995), then hearing captioning viewers are 865 times as common as deaf or hard-of-hearing.
So, to use numbers of decoders rather than users, if slightly more than 1/532 or 0.1879% of those people turn captions on, hearing people outnumber deaf as caption viewers. To use Kaye’s numbers, if slightly more than 1/865 or 0.1156% of those people turn captions on, hearing people outnumber deaf as caption viewers.

Canadian calculations

Industry estimates of TV sales in Canada are as follows:

1993: 1,511,000 (we’ll be using half this number: The decoder law was not in effect for the first half of 1993)
1994: 1,545,000
1995 and later (my annual estimates): 1,540,000

Accordingly:

In Canada, a total of 12,296,000 TV sets were sold from mid-1993 to 2000.
Assume 90% of those sets are 13 inches in size or larger. That means 10,386,450 decoder-equipped TV sets were sold since the decoder law kicked in, which is 667 times the number of set-top decoders sold in Canada.
If 90% of those sets are watched by hearing people (a conservative estimate), then 9,347,805 decoder-equipped TV sets were sold since the decoder law kicked in, which is 600 times the number of set-top decoders in Canada.
So if slightly more than 1/600 or 0.1666% of those people turn captions on, hearing people outnumber deaf as CC viewers.

Conclusions

In the year 2001, all it takes are two hearing viewers out of 1,000 to turn on captions in order for hearing people to become the majority audience of captioning. If that number nonetheless seems too high, just wait a few years. With 21,497,400 decoder-equipped sets bought by hearing people annually in North America, every passing year puts a big dent in the percentage of hearing people who need to watch captions in order for hearing people to become the majority audience. (21.5 million decoder-equipped sets bought by hearing people amount to 72 times as many set-top decoders as were ever sold.)

This analysis does not indicate that captions should be optimized for hearing people – in other words, with no rendition of dramatically significant sound effects or tones of voice and no positioning or explicit identification of speakers. Captions must continue to be optimized for an assumed deaf audience; it is merely a consequence of the ubiquity of built-in decoders that hearing people are now the largest audience.

Limitations

This analysis does not use super-exact numbers. But since the source figures hover in the millions, the number of significant digits would be unaffected by pinpoint precision.
It does not attempt to estimate the number of set-top decoders sold in the present day (it is still possible to buy one, like the MyCap Junior).
Nor does it estimate the tiny sales figures of the so-called Integrated Television Receiver, a TV-plus-decoder combination sold by Sears in the very, very earliest days of captioning.
Further, Gary Robson reports that Zenith (and a few miscellaneous vendors) sold a million decoder-equipped TVs in the days before mandatory decoders. These numbers are not included here, but they would not change the basic thrust of the observed effect.
Statistics Canada’s estimates of decoder use are old and unreliable, but they’re all we’ve got. The 1994 data quoted above actually comes from a 1991 survey, which is now a good ten years old. A separate survey, “Canadians with Impaired Hearing” (1992; catalogue 82-615; ISSN 1180-4610), working with data from 1987, states that 7,570 people aged 15 or older used caption decoders. But 1987 was pretty early in the development and acceptance of captioning in Canada; the low number is to be expected.
Among hearing people, the calculations do not differentiate second-language learners from English native speakers, nor does it separate French-language viewers from English-language in Canada.
Finally, I have no citation for the figure of 300,000 set-top decoders sold. (The citation is lost. I hope to eventually recover it.) The Caption Center arrives at a slightly larger estimate: 350,000. Taken to the extreme, however, even if there were three million set-top decoders in use, this analysis would not be hugely altered.