Importantly, the goal here was to estimate the effective value of each radio spin *per listener* so that rate can be compared to not only that rate being observed for the effective value of a stream on music services (e.g., Rdio, Rhapsody, or Spotify) per listener, but also those rates applicable to webcasting streams.

First, a quick clarification: Most chatter about radio royalties hinges on the value of a spin, and yet the chatter about music service royalties hinges on the value of a stream. However, this comparison between spins and streams is not usually/always equivalent. Why?

The usual “spin” value often refers to a radio play on a station or syndicated show that was experienced by a large audience (sometimes millions of people).

A stream payment usually refers to a play/stream by a music service experienced by a single listener (or at least a single subscriber account).

Therefore, I am trying to compare spins and streams on equal terms by converting this spin-to-millions value from Radio to a value per listener, as is the case for music services. Please scroll to the end of this post for an explanation of the per spin/stream per listener comparison in the context of licenses based upon percentages of revenue or blanket fees and advances.

Now for the numbers…

The Short Story:

Just the calculated rates and none of the overly numerical explanations…

Important: effective rates below are estimates of amounts paid only for the performance of the musical work (song/lyrics) and to the stakeholders of those works (publishers, songwriters, and lyricists).  In the US, radio stations pay royalties for only these performance rights and do not pay for performing rights to the stakeholders of the sound recording.

Taking a stand:

$0.0002977923 <- Estimated effective rate per spin per listener*
$0.0000992641 to $0.0005955847 <-Range**
at 15 songs per hour

$0.0003722404 <- Estimated effective rate per spin per listener*
$0.0001240801 to $0.0007444808 <-Range**
at 12 songs per hour

$0.0004963206 <- Estimated effective rate per spin per listener*
$0.0001654402 to $0.0009926411 <-Range**
at 9 songs per hour

$0.0013 <- Effective rate per stream per listener on music services
$0.000300 to $0.0015 <- Range
estimated through songwriter reports, musical work only

The Long Story (with the Inputs):

The effective rates paid by radio and music services in the US are less directly comparable to those rates paid in the UK. Unlike in the UK, radio broadcasters in the USA pay royalties to only one set of music rightsholders: those with an interest in the musical work [the publisher(s), lyricist(s), and songwriter(s)]. Music services in both countries, on the other hand, are paying both sets of rightsholders.

I will begin with the total Dollars collected from Radio by the three performance rights collectives in the US: ASCAP, BMI, and SESAC. While ASCAP openly discloses its domestic collections from Radio, BMI only discloses these values indirectly (in its annual report)—reporting the proportion of total collections that are not international, and that proportion of domestic collections that are from Radio. I had to estimate SESAC’s collections. Being a private company, SESAC does not publicly disclose its collections.

I will then break down this total value of collected Dollars on a daily, hourly, and finally per spin basis (for the set of music tracks hitting the ears of listeners at any moment in time in the US across all music radio stations).

Since not all radio stations are playing music (predominantly or exclusively), I have chosen to follow the royalty dollars to the radio-listening audience, through the tracks plausibly heard by only this audience.

Finally, I arrive at an effective range of the per spin per listener value through a set of estimates of the average size of only the music radio listening audience during the day in the US.

The overly numerical banter:

(A) $230,881,000 <- Dollars collected from Radio by ASCAP (pdf)

(B) $198,303,000 <- Dollars collected from Radio by BMI (imputed)

(C) $42,918,400   <- Dollars collected from Radio by SESAC (estimated)

(A)+(B)+(C)          <- Dollars collected from Radio by the Collectives
equals
$472,102,400       <- Estimate of Total Dollars collected from Radio in the US

/ 365 <- Number of days in the year, resulting in per day payments
/ 24   <- Number of hours in the day, resulting in per hour payments
equals
$53,892.97 <- Dollars collected per hour from US Music Radio

So, if music royalties were measured like a taxi meter, the meter would rack up royalties at a rate $53,892.97 per hour.

/ 12 <- Number of songs played per hour, resulting in per song payments
equals
$4,491.08 <- Dollars collected per simultaneous spins across entire US Music Radio

The meaning of this $4491.08 is simple.  PROs collect royalties from the performance of about X songs per hour (where X might be 12, or 15, or 9) across all music radio stations.  $4,491 is simply the Collections per Hour divided by X, if X were 12 songs per hour.

Note: That figure—$4,491.08—is not the value of a spin on a single station or show, like Jack FM. It is an estimate of the average total value of all spins across all music stations in the US given some number of songs played per hour.  A source like Jack FM would be one of the shows/stations playing music at any time. If, however, I reckoned that any station/network were 10% of the listening audience at some point in time, the value of that spin on that single station might be around $449.11.

Some might disagree with the rate of 12 tracks per hour (or about 36 minutes of music). And so I also ran estimates asserting 15 tracks per hour (about 45 minutes of music) and 9 tracks per hour (24 minutes of music).

We need to divide $4,491.08, however, by the average size of the US radio audience throughout the day that is listening to music radio to get an estimate of the value of each spin per listener. Essentially, if I could flip a switch and know how many people were listening to music on the radio at any moment, how many people (on average) would I find to be listening?

Unfortunately, such a switch does not exist. But a variation exists in the form of Arbitron’s estimates of the radio audience.

In the US, Arbitron estimated that during the most recent quarter approximately 241,300,000 people over the age of 12 listened to the radio at some time. This estimate is not the same as an estimate of the average size of the listening audience, however. In particular, it is not that portion of the audience that is listening to music (on the) radio at moments in time.

Arbitron estimates Hour-by-Hour listening as a proportion of its quarterly radio audience. In their 2010 Radio Today report (page 89), these ratings throughout the day range from a high of approximately 17% of quarterly listeners to a low of around 1% of quarterly listeners. The average across the 24-hour period appears to be in vicinity of 9.5%.

241,300,000 <- Arbitron estimate of quarterly radio listening audience, Q3 2011

$4,491.08      <- Dollars collected per spin if across entire US Radio audience

/ 6,032,500   <- approximately 2.5% of the quarterly radio audience
or
/ 12,065,000 <- approximately 5% of the quarterly radio audience
or
/ 24,130,000 <- approximately 10% of the quarterly radio audience
or
/ 36,195,000 <- approximately 15% of the quarterly radio audience

—————————-

Dollars collected per spin per listener for Radio in the US for various estimates of the average listening audience:

At 15 songs per hour
$0.0005955847 <- if the average listening audience is 6,032,500 people
$0.0002977923 <- if the average listening audience is 12,065,000 people
$0.0001488962 <- if the average listening audience is 24,130,000 people
$0.0000992641 <- if the average listening audience is 36,195,000 people

At 12 songs per hour
$0.0007444808 <- if the average listening audience is 6,032,500 people
$0.0003722404 <- if the average listening audience is 12,065,000 people
$0.0001861202 <- if the average listening audience is 24,130,000 people
$0.0001240801 <- if the average listening audience is 36,195,000 people

At 9 songs per hour
$0.0009926411 <- if the average listening audience is 6,032,500 people
$0.0004963206 <- if the average listening audience is 12,065,000 people
$0.0002481603 <- if the average listening audience is 24,130,000 people
$0.0001654402 <- if the average listening audience is 36,195,000 people

Finally, and importantly, I realize that Radio stations in the US do not pay royalties on a per spin (and per listener) basis. Instead, these Radio royalties are calculated by such means as a percentage of revenue, or as a result of a negotiation over a blanket fee.

I also realize that music services—such as MOG, Rdio, Rhapsody, and Spotify—are not—by default—paying royalties on a per stream (per listener) basis. Instead, these royalties may be calculated through a number of means, none of which can be openly discussed due to the joys of non-disclosure clauses.

However, the resulting pool of money from Radio can be and has been counted. And various sources have reported online their payments from music services. As a result, nerds like me can try to estimate some effective rate given the size of the pools, the rough number of tracks played, and the number of listeners.

*
The estimates above are a function of considering the average music radio listening audience (across moments in the day) to be 5% of the estimated quarterly radio audience as reported by Arbitron. For comparison, Arbitron estimates the average size of US Radio (all formats) audience to be approximately 9.5% of the quarterly audience. I worked below this number given (a) not all stations are music stations and (b) Arbitron may be overestimating the size of the Radio audience.

**
The ranges above are a function of considering the average music radio listening audience (across moments in the day) to be 2.5%, 5%, 10%, and 15% of the estimated quarterly radio audience as reported by Arbitron. For clarity, the smallest assumed listening audience will lead to the highest estimated rate per stream per listener in the ranges.

Below I try to calculate the effective value of a radio spin *per listener* so that rate can be compared to the rate being observed for the effective value of a music service stream per listener.

First, a quick clarification: Most chatter about radio royalties hinges on the value of a spin, and yet the chatter about music service royalties hinges on the value of a stream. However, this comparison between spins and streams is not usually/always equivalent. Why?

The “spin” value often refers to a radio play experienced by a large audience (sometimes millions of people).

A stream payment usually refers to a play/stream by a music service experienced by a single listener (or at least a single subscriber account).

Therefore, I am trying to convert this spin-to-millions value from Radio to a value per listener, as is the case for music services. Please scroll to the end of this post for an explanation of the per spin/stream per listener comparison in the context of licenses based upon percentages of revenue or blanket fees and advances.

Warning: I end up with a rather small number as the estimate of the value of a UK Radio spin per listener. Fortunately, and importantly, the radio audience in the UK is quite large overall, leading to a total pool of royalty monies that is large enough for most labels, artists, songwriters, and publishers to take notice.

Were the pool of money flowing from music services in the UK, such as Spotify, as large as that flowing from Radio, we might not be having these sorts of discussions. Maybe someday both of these pools will be large. The bigger question than these present rates, therefore, is whether we should and how we might get to that version of someday.

The Short Story:

just the calculated rates and none of the overly numerical banter…

£0.0002992992 <- Estimated effective rate per spin per listener
£0.0000998 to £0.0005985 <-Range
at 15 songs per hour

£0.0003741240 <- Estimated effective rate per spin per listener
£0.0001247 to £0.0007482 <-Range
at 12 songs per hour

£0.0004988320 <- Estimated effective rate per spin per listener
£0.0001662 to £0.0009976 <-Range
at 9 songs per hour

The estimates above are a function of considering the average music radio listening audience (across moments in the day) to be 5% of the estimated quarterly radio audience as reported by RAJAR. The ranges above are a function of considering the average music radio listening audience (across moments in the day) to be 2.5%, 5%, 10%, and 15% of the estimated quarterly radio audience as reported by RAJAR. For clarity, the smallest assumed listening audience will lead to the highest estimated rate per stream per listener in the ranges.

£0.0005000 to £0.005000 <- Music Services effective rate per stream per listener (estimated)

The Long Story (with the Inputs):

The effective rates paid by radio and music services in the UK are more directly comparable as compared—no pun intended— to those rates paid in the US. Unlike in the US, radio broadcasters in the UK pay royalties for both sets of music rightsholders: (a) those with an interest in the sound recording [aka, the neighbouring rights; the owner(s) of the recording, as well as the featured and performing artist(s)] and (b) those with an interest in the musical work [the publisher(s), lyricist(s), and songwriter(s)]. Music services are paying both sets of rightsholders as well.

I will begin with the total Pounds collected from Radio by the two core performance/performing rights collectives in the UK: PRS for Music and PPL.

I will then break down this total value of collected Pounds per day, per hour, and finally per spin (for the set of music tracks hitting the ears of listeners at any moment in time in the UK across all radio stations).

Since not all radio stations are playing music (predominantly or exclusively), I have chosen to follow the royalty dollars to the radio-listening audience, through the tracks plausibly heard by only this audience.

Finally, I arrive at an effective range of the per spin per listener value through a set of estimates of the average size of only the music radio listening audience in the UK during the day.

The overly numerical banter:

(A) £49,500,000 <- Pounds (2010) collected from Radio by PRS (pdf)

£61,700,000 <- Pounds (2010) collected from Broadcast/Online by PPL
x 70% <- Rough approximation of % of PPL B/O from Radio**
equals
(B) £43,190,000 <- Estimate of Pounds collected for Master Performance

(A) + (B) <- Pounds collected from Radio by PPL and PRS
equals
£92,690,000 <- Estimate of Total Pounds collected from Radio in the UK

/ 365 <- Number of days in the year, resulting in per day payments
/ 24 <- Number of hours in the day, resulting in per hour payments
equals
£10,581.05 <- Pounds collected per hour from UK Music Radio

So, if music royalties were measured like a taxi meter, the meter would rack up royalties at a rate £10,581.05 per hour.

/ 15 <- Number of songs played per hour, resulting in per song payments
equals
£705.40 <- Pounds collected per spin across entire UK Music Radio

Note: That figure—£705.40—is not the value of a spin on a single station, like BBC Radio 1. It is an estimate of the average value of all spins across all music stations in the UK at a moment in time, one station of which would be the BBC’s Radio 1. If, however, I reckoned that BBC Radio 1 were 10% of the listening audience at some point in time, the value of that spin on that single station might be around £70.54.

Some might disagree with the rate of 15 tracks per hour (or about 45 minutes of music). And so I also ran estimates asserting 12 tracks per hour (about 36 minutes of music) and 9 tracks per hour (24 minutes of music).

We need to divide £705.40, however, by the average size of the UK radio audience throughout the day that is listening to music radio to get an estimate of the value of each spin per listener. Essentially, if I could flip a switch and know how many people were listening to music on the radio at any moment, how many people (on average) would I find to be listening?

Unfortunately, such a switch does not exist. And so, now is that uncomfortable time when assumptions, guesstimates, and hopefully well-reasoned estimates are made.

In the UK, RAJAR estimated that during the most recent quarter approximately 47,137,000 people over the age of 15 listened to the radio at some time. This estimate is not the same as an estimate of the average size of the listening audience, however. In particular, it is not that portion of the audience that is listening to music (on the) radio at moments in time.

Below, I will calculate an estimate of the value per spin per listener at three guesstimates of the average size of the music radio listening audience during the day in the UK, given the peak/trough variation in this audience. Please feel free to share your thoughts/insights as to what other inputs I could/should use here.

Added: In the US, Arbitron estimates Hour-by-Hour listening as proportion of its quarterly radio audience. In their 2010 Radio Today report (page 89), these ratings throughout the day range from a high of approximately 17% of quarterly listeners to a low of around 1% of quarterly listeners. The average across the 24-hour period appears to be in the 8-10% range.

47,137,000 <- RAJAR estimate of quarterly radio listening audience, Q3 2011

£705.40 <- Pounds collected per spin if across entire UK Radio audience
/ 2,356,850 <- approximately 5% of the quarterly radio audience
/ 4,713,700 <- approximately 10% of the quarterly radio audience
/ 7,070,550 <- approximately 15% of the quarterly radio audience

—————————-

Pounds paid per spin per listener for Radio in the UK for various estimates of the average listening audience:

At 15 songs per hour
£0.0002993 <- if the average listening audience is 2,356,850 people
£0.0001496 <- if the average listening audience is 4,713,700 people
£0.0000998 <- if the average listening audience is 7,070,550 people

At 12 songs per hour
£0.0003741 <- if the average listening audience is 2,356,850 people
£0.0001870 <- if the average listening audience is 4,713,700 people
£0.0001247 <- if the average listening audience is 7,070,550 people

At 9 songs per hour
£0.0004988 <- if the average listening audience is 2,356,850 people
£0.0002494 <- if the average listening audience is 4,713,700 people
£0.0001662 <- if the average listening audience is 7,070,550 people

Finally, and importantly, I realize (realise) that Radio stations in the UK do not pay royalties on a per spin (and per listener) basis. Instead, these Radio royalties are calculated by such means as a percentage of revenue, or as a result of a negotiation over a blanket fee.

I also realize (realise) that music services, such as Spotify, are not—by default—paying royalties on a per stream (per listener) basis. Instead, these royalties may be calculated through a number of means, none of which can be openly discussed due to the joys of non-disclosure clauses.

However, the resulting pool of money from Radio can be and has been counted. And various sources have reported online their payments from music services such as Spotify. As a result, nerds like me can try to estimate some effective rate given the size of the pools, the rough number of tracks played, and the number of listeners.

** As far as I could tell, PPL does not breakdown its Broadcast and Online revenue further in its Annual Report. There was not direct comparable for this estimate of the proportion of PPL collections that are derived from UK Radio royalties. I erred on the side of too large a proportion rather than too small. Any guidance on this figure would be appreciated

I figured I would respond to some of the comments and criticisms, however.  Frankly, I appreciate both.

Your estimate of Spotify’s pay rate is off, no one is getting $0.0033 per stream:

I employed only the effective rates that had been posted publicly by a variety of sources. I was trying to work with an effective rate per stream (recognizing that deals may not be per stream rates, necessarily).

Those sources:

PaidContent reported that STHoldings claimed $2500 from 750,000 plays, which works out to about $0.00373 per stream. ST Holdings removed the post with this claims (I believe) as well as their content from Spotify for the reasons of the payout levels.

Gizmodo published a post from Uniform Motion who reported receiving $0.0029 Euros per stream ($0.00386 at present exchange rate or $0.0041 by Gozmodo’s calculations).

Billboard.biz reported that Mode records calculated receiving $36.98 for 11,335 stream = $0.0033 per stream.

Digital Audio Insider reports average payout of $0.002865 per stream.

DigiMuziek offered up screenshots of their payments, which were increasing from EU0.0019 to EU0.0022 from September to October 2011. The last rate works to about $0.0029 per stream.

Some Artists/Labels in the US are reporting lower amounts.  These per stream rates seem to be clocking in closer to the $0.0013 per stream that Sam from Projeckt Records is reporting.  I am looking into these lower rates to try to understand them and the differences.

Let’s be clear, however.  Different people appear to be receiving different effective rates.  That difference matters.

Spotify does not really pay a set per stream rate:

OK.  But I was trying to compare an effective rate per listen for both downloads and streams.  To do this I had to estimate something akin to a per stream rate.  Your contract may not say “per stream rate.” There still would exist an effective rate per stream (dollars received divided by number of streams).

Spotify (unlike some other music services) would like to operate by paying artists a percentage of revenue rather than a set rate per stream.  I tried to do the same thing with a music service more than a decade ago.  This sort of structure “makes sense” from the point of view of the service.  It may not make the same sort of sense for rightsholders.

When the rubber hit(s) the road, I think its worth debating whether certain rightsholders are seeing guarantees or minimas that present services such as Spotify with essentially fixed-rates in some cases, fixed portions of subscriber fees in other cases, and just raw % of revenue in others.

250 listens to a track over the lifetime is too high:

I may have overestimated that.  But to be clear, I was comparing purchased track downloads to music service streams.

The feedback I got was that the average is quite different if you look at single track downloads you purchased (higher range average listens), albums you purchased that have tracks you never wanted in the first place (middle range average listens), and giant music collections the bulk of which you simply obtained and maybe did not acquire (low range average listens).

Comparing CD sales to Music Services streams:

Some people responded by comparing CD sales to Spotify streams.  CDs have ten or more tracks on them.  So a single CD sale is actually 10 (or more) track sales.

CD/Track sales equate to millions of Music Service streams:

You will unquestionably find that CD sales in the near term are equivalent in dollar terms to millions (if not tens of millions) of streams on a music service.

I wasn’t denying this difference.  The different was the point of the analysis and estimates.

I was trying to understand whether these amounts were that different if you considered them in equivalent terms—payments per listen.

To be continued…

And, first off, many thanks to Jeff and Tunecore for releasing these data for the rest of us to investigate.

It appears life in the long tail of music sales may be rather rough. DMN has one angle (minimum wage), while I took a sales distribution approach.

Of the top 6000-ish (5938) artists about which Tunecore just released sales data, the sales distribution follows an 80/20 rule relatively closely. Collectively, the top 22% of these artists earned about 80% of the monthly money.

Given these 6000 are only 1% of reported 600,000 accounts, however, it would appear that Tunecore artist sales exhibit an extremely pin-headed distribution rather than a long tail distribution (as per Chris Anderson).

In fact, these 6000-ish artists produce enough revenue that, if converted to annual dollars, they might be earning half or more of total Tunecore artist revenue.

Since Spotify is really just one of the music services alive and kicking on the planet these days (alongside Rhapsody, MOG, Rdio, Deezer, and others), its worth investigating this matter of the value of payments from streaming services versus those received from record sales.  So let’s get a napkin out and flip to the backside for some mathy stuff.

Assertion:

When you purchase a download you are paying in advance for all of your subsequent listens of that track.

When you listen to a track through a streaming music service, you are paying as you go for each and subsequent listens of that track.

First off, its important that we compare apples to apples and not apple to oranges. Comparing the value of a stream to the value of a sale is without question an Apples to Oranges comparison.  And so, let’s convert the sale to an estimate of payment per stream so we can compare apples to apples (or oranges to oranges if you prefer).

Second, let’s make a simple assumption: A single track purchased from iTunes will be listened to 250 times, on average, over the lifetime of the person who acquired the single.  This assumption may not seem that crazy—over twenty five years will you listen to the “average” purchased song in your library about 10 times per year?  Or twenty five times per year over the first ten years?

Dividing the amount paid for a single by the number of times you listen to that single over the lifetime will give us the effective rate you have paid for each listen of the track.  This payment per listen is the closest equivalent to the payment per stream courtesy a music streaming service.

Third, now we can just divide the amount music rightsholders would earn from a sale on iTunes to estimate a “per listen” fee—the effective rate paid over the lifetime of the track’s enjoyment for each listen of the song.

$0.70 -> the amount a label receives for the sale of an iTunes single (US)
250    -> the number of assumed listens to the track over ownership life
——————–
$0.70 / 250 =  $0.0028 per listen

OK.  So some quick math has led us to the conclusion that our “per listen” payment for iTunes singles over the lifetime of our enjoyment of these tracks is just short of 1/3 of a penny.

From the perspective of songwriters and publishers, in the US, some of this $0.70 above should be flowing back to you—about 9.5 cents (give or take for those various alterations like controlled compositions). And so, the per listen portion paid to the songwriter/publisher would be around $0.00038, leaving about $0.00242 for the sound recording owner(s) and performing/featured artist(s).

To repeat, if you convert purchases of downloadable singles to some value “per listen,” we get the following:

$0.00242  -> for the label(s) and performing artist(s)
$0.00038 -> for the songwriter(s) and publisher(s)
——————-
$0.0028   -> total amount paid to rightsholders per listen from a purchased track

Now we can ask, Is $0.0028 greater than or less than what rightsholders are receiving per stream from streaming music services?

Since disclosing the licensed rate at which these music services pay for music per stream would be controversial behavior, let’s just use the going rate as suggested by those who have chosen to disclose (or claimed to disclose) their payments from Spotify.  That rate is apparently about $0.0033 per stream paid to the label, some portion of which may or may not ever make its way back to the artist(s).

$0.0033 -> the amount believed to be paid per stream via Spotify (for SR only)
$0.0028 -> the effective “per listen” payment for a purchased download
——————–

$0.0033 > $0.0028

Could it be true?

Is the per stream rate paid by music services actually greater than the effective per listen rate we pay for a purchased download?

Next up, we will use some not-so-fancy financial tools (NPV) to compare the present value of these two payments schemes: up-front versus over-time.

Spotify free is free and unlimited in usage for only the first six months.  After that term, users will be limited to ten (10) hours of free listening in any month, with a cap of five (5) listens to any particular track.  Unclear whether the # of listens limitation renews each month or is a true cap.

The quoted language of the Spotify TOS follow:

The Spotify Service can be accessed

(i) as an ad-supported free-to-the-user service having no monthly cap on listening hours or a cap on number of plays of a unique track during the first 6 months following creation of your Spotify account but thereafter a cap of 10 listening hours per month and a cap of 5 plays per unique track (the “Free Service”),

(ii) as an advertisement free version of the Spotify Service for which you pay a monthly or yearly subscription (the “Unlimited Service”),

(iii) as an advertisement free version of the Spotify Service including an offline mode which makes it possible to listen to music without a connection to the Internet via conditional downloads and other additional features for which you pay a monthly or yearly subscription (the “Premium Service”), and

(iv) as a version of the Spotify Service that provides access to aspects of the Spotify Service via a supported mobile handset (the “Mobile Service”).

Always read the fine print.

A “lean back” version of the presentation is below. Sorry, I haven’t had time to do a voiceover.

A number of people have asked for a version of the presentation. Please contact me directly.

Anyone who has completed and defended their dissertation knows that the best dissertation is the done dissertation.  My dissertation is now done as in defended, approved by the committee with all committee member signatures, submitted to the graduate school, approved by the graduate school, and the confirmation email from the graduate school stating that I have completed all the requirements for a PhD in Management has been received.

While the dissertation defense is an altogether unique experience, I reckon the best way to experience a dissertation defense presentation is in fast-forward (rather than in real-time).  Therefore, I have placed my defense presentation online in a video rolling at the rate of 100 slides in 2 minutes and 45 seconds.  So hit play and then sit back and watch a few years of work fly by in less than three minutes.

Ghosts in the Shell: An investigation of the relationship between automation and the nature of work from david touve on Vimeo.