Bottom Line With the rise of over-the-top (OTT) services offered from non-traditional providers like Netflix, download services like iTunes 256 application program , clips from services like YouTube, and the continued usage of pirate sites (streamed or downloaded), talk of the imminent collapse of traditional TV is understandable. But while the US TV market is not growing, it is not collapsing either. The best way of describing the outlook is gradual erosion: an apocalypse is not around the corner. An obvious question is: “What do all these various erosions do to the size of the US television industry?” If it is about $170 billion in 2016, (of which $75 billion is advertising and $95 billion is pay-TV) how does that number change thereafter, and by how much? The outcome is hard to predict with certainty. Putting it all together, the total picture is murky. If one percent of current pay-TV subscribers discontinues service in 2016 this does not necessarily mean a one percent decline in revenues. Cord cutters are likely to be those on the lowest-priced pay-TV packages. Their departure may cause ARPU to rise and 257 become more resilient: customers that remain may be those least sensitive to further price increases . Cord cutters are likely to include those watching the fewest minutes of TV, and may represent a smaller share of ad dollars to the industry than subscription dollars. Some cord cutters are moving to OTA antennas, and may therefore end up watching more ads due to the lack of a DVR, and subscription losses may be offset by more effective advertising. Equally, cord shavers who get rid of channels they are not watching will likely have minimal impact on ad revenues. Millennials aged 18-24 are a desirable demographic, and they are reducing TV minutes at a faster rate than the population as a whole, 258 but are only a tenth of the US population . Further, advertisers often pay to target specific groups: if traditional TV watching becomes concentrated in certain age groups or other demographic slices, advertisers could be willing to pay more for that targeted audience. As an example, there were concerns over weakness in the TV ad market in the summer of 2015, as the mid-point of the August 259 260 US upfront estimates fell over two percent from 2014 . However by October, total TV ad spend was up 10 percent annually . The US is only one market, albeit the largest in the world, representing about 38 percent of the global TV market of around $450 billion worth of subscriptions, advertising and license fees. What happens in the US may or may not happen in the rest of the world. Trends in Canada for cord cutting, cord shaving, pay-TV penetration and changes in viewing for the population as a whole, and for millennials, are roughly in line with the US data cited. If we look at the UK TV market there are some similarities: viewing minutes are expected to fall between 2015 and 2020, but 261 moderately, from 204 daily minutes to 191 . The UK data also shows a 27 percent decline in viewing by 16-24 year-olds for the period 2010-H1 2015: from 169 daily live and time-shifted minutes to 123, very much in line with the US decline of 29 percent over 262 (roughly) the same period . However, the forecast is for the peak decline in that youngest demographic to have occurred in 2014/15, 263 and to “flatten out from 2016 onwards” . It is also worth noting that although TV minutes may be declining in the UK, the pay-TV 264 market is growing strongly, both in terms of subscribers and revenues , and the UK TV advertising market grew eight percent in 265 2015, recording its best growth in 20 years . Although pay-TV subscribers are expected to fall slowly in the US, the global picture remains in growth mode. From 950 million subscribers and 58 percent penetration in 2015, estimates for 2020 are for 20 percent growth to 1.14 billion subscribers and 266 63 percent penetration . Our thesis of gradual erosion is based on recent history and incomplete data. As already mentioned, if there is a tipping point of viewing hours needed to sustain pay-TV subscription rates, we could see the number of millennials who have their own homes and 267 who do not get pay-TV go from current levels of 20-25 percent to a much higher number in a short period of time. That hasn’t happened yet, but it could. Another wild card is likely to be the effect of popular channels that had formerly been available only as part of a pay-TV bundle being offered on a stand-alone basis over the top (OTT) through the Internet. One such service has already launched 268 in the US, and has only seen about one percent of its pay-TV subscribers cancel pay-TV and switch to the OTT version . Should that service, or any others, see that substitution accelerate, the fragmenting of the traditional pay-TV bundle model will almost certainly see our predictions on cord cutting, cord shaving, and cord nevers prove too conservative. Finally, although the 18-24 year-old category is currently moving the fastest away from the traditional TV model, it is worth adding that older generations are also exhibiting some of the same shifts, albeit at lower levels. If those older age groups began to resemble millennials more rapidly in their pay-TV habits, our forecasts would again be too cautious. 38