On Twitter: @JamesFirth and @s_r_o_c (post feed)

Got a tip? tip@sroc.eu

Thursday, 30 May 2013

Sources: no Digital Economy Act copyright warning letters until 2016 at the earliest

It's been a while since I wrote anything on the Digital Economy Act.

As a quick recap, although the primary legislation to notify and punish those whose internet connection is repeatedly used to infringe copyright was rushed through the dog end of the Labour government - with Conservative support from the then shadow Culture Secretary Jeremy Hunt - progress has since stalled.

Two Statutory Instruments were expected to flesh-out how the copyright infringement warning letters will be dispatched, paid for and appealed: a so-called Initial Obligations Code and a shorter Cost Sharing Order defining how the cost of scheme will be split between copyright owners and internet service providers.

Two attempts have so-far been made to enact the shorter Cost Sharing Order, but even this relatively simple piece of legislation only got as far as the notification phase, where other EC member states are notified of draft changes to policy potentially affecting cross-border trade [refs: here, and here].

Both times the Order was quietly dropped, and nothing much has been heard since.

In response to a Tweeted question this morning I thought I'd see if my old Westminster contacts still wanted to talk to me.

Two separate sources told me not to expect the remaining secondary legislation this side of the general election.

Assuming a 2015 general election, and factoring-in time to establish the necessary body or bodies to oversee the operation of the notification and appeals systems, it will be 2016 at the very earliest - and possibly 2017 - before the first warning letters go out.

UPDATE 19-June

5 days after I posted this, DCMS released minutes [link] outlining their expectations:
"DCMS expects the first letters to be sent in the latter half of 2015"
My own sources were clear that the remaining legislation - two statutory instruments (that might possibly, now, be rolled into a single instrument) - were unlikely to go before Parliament until after the 2015 General Election.  By my own estimates it will then take around 9 months to establish the systems necessary to get the first letter out, including the crucial appeals process, hence my claim of 2016.

It seems the Department for Media, Culture and Sport (DCMS) thinks it can be done slightly faster, hence their note of optimism in their timetable of back-end of 2015.

So what's the delay?

One source described the copyright provisions in the Digital Economy Act 2010 as "un-implementable".

The legislation rushed through parliament in 2010 - at the behest of copyright lobbyists asserting prompt action was essential to the survival of the creative industries - was bad.

Since then the UK music and film industries have grown despite the gloom in the rest of the economy and 2012 saw an 11% revenue growth for legitimate downloaded media content in the UK despite progress on the Digital Economy Act stalling.

Tuesday, 28 May 2013

Do established economic models apply to data, and will they spread wealth, or is Jaron Lanier wrong?

Jaron Lanier's article on the BBC is really worth reading.  His is the first mainstream article in a while that has got me wrestling with my instincts on big data, privacy and digital democracy.

I agree with a many of his observations.

Yet I disagree with his primary thrust, that "open" has backfired and instead created a new data inequity; and, controversially:
"monetising information will bring benefits that far outweigh the inconvenience of having to adjust one's worldview."
That there is a new power "asymmetry" is undoubtedly true.  I wrote about the new data plutocrats just a fortnight ago.

But would monetisation of our data have worked any better?  

I argue not.  I believe the largely cashless and open information society emerging over the last two decades was a necessity in order to overcome intransigence and challenge the existing economic power balance; and that, had our data been monetised along the lines suggested by Lanier, we'd be far worse-off today.

Had data been monetised, the primary beneficiaries would have been those who currently control the money supply; "data" power would be controlled by those who already wield a great deal of economic power, and democracy would have suffered.

Essentially, Google and gang have emerged, warts and all, to counter the dominance of not just the more established technology companies but also the big banks, and even governments.

Not that I don't foresee severe problems ahead with the path we're now travelling, most of these problems adequately covered in his article.

But, for the time being at least, the old economy tempers Google's power just as Google threatens the old economy.

Inherent constraints of cash

Perhaps, even more importantly, monetising data might have constrained innovation because there is not [yet?] an effective mechanism for setting a price on such a rich and nuanced commodity.

What's a Tweet worth? How much should I be paid, or pay, to participate daily in Facebook? Is my social capital worth anything?.. To me?.. To you?

Such questions may prove unanswerable because the concept of money a hundred millennia old has its limitations.

It may simply prove impossible to create a market that accounts for the diversity of the information economy and its many applications: from furtherance of knowledge to private and personal sharing to providing transparency of governments.

And attempting to do so in a flawed way may derail rewarding but otherwise economically non-viable data services.

For it is possible for us to be getting monetarily poorer and richer at the same time when our lives are getting richer in other ways, even as the 'real' economy stagnates.

Data, and with it, cheap access to communication tools, undoubtedly enrich our lives, and this is both the reason data is hard to monetise and the foundation of the data bartering Larnier sneeringly dismisses as "usually associated with the developing world."

Instead of trying to shoehorn data into GDP we should be looking to an evolution of money to measure our overall wealth, and also to restore equity to contributors and remunerate "workers" - something Lanier correctly notes as absent today.

A Necessity: the Open sledgehammer

As well as overlooking the limitations of money, Jaron Lanier fails to address the positive and necessary role the Open movement played in challenging the inertia present in any stable society.

When I started my career in software in the mid 90's there were sizeable barriers to entry.

Firstly, the software tool chain was largely closed, and licencing costs were huge.  A handful of then-dominant global corporations controlled access to most platforms. (In a way that Apple does, albeit with much lower barriers to entry for developers, today.)

To get [legal] access to a software compiler and the rest of the development kit (SDK) companies had to fork out thousands.  Add to that: revision control software, a defect tracking database, operating system licenses for servers and worksations, etc, and you were talking considerable start-up costs.

It was a system that actively prevented the kind of "bedroom innovation" that has created revolutionary apps and services in the last two decades.

In fact, for a while at least, innovation throughout the sector seemed to stall as larger tech companies, in the absence of serious competition, took their foot off the gas in order to consolidate.

Some large corporations placed emphasis on pursuing licensing revenue for existing products above investing in product innovation, choosing to pursue patent and other IPR infringement at the expense of developing better software.

It really was a frustrating time - at least through the eyes of a twenty-something-year-old.

Developers were left with buggy, feature-void tools.

Smaller companies in particular hesitated to buy the latest software.

And suddenly, innovation was threatened on every front: big companies wanted to return a dividend to their investors, smaller companies were stuck doing things the hard way and tiny companies couldn't even afford to enter the market.

Around that time many in my position had been exposed to Linux and the Open Source movement at schools and universities (Slackware for me, circa 1994).

Before long I could build bigger and better software on my home-built server (primarily around Apache and Perl or related CGI scripting language) than I could using the licensed tool-set my company paid for at work.

In asserting that we may have denied ourselves something even better by turning our back on monetisation overlooks how and why we are where we are now and  fails to acknowledge necessity - the driving force for many open source projects.


Friday, 17 May 2013

Time for a new plan for corporation tax to re-level the field for local employers

What keeps me mad throughout the ongoing corporation tax bunfight is that I agree with no-one.

On one [typically right] side, there is the argument that companies such as Amazon, Vodafone, Starbucks, Google, etc, etc... should not be criticised because they're only doing what they are obliged to do: minimise their tax bill.

Fair enough, but we live in a democracy, so claiming they shouldn't be criticised or the subject of peaceful protest is a bit far fetched.  If enough people feel aggrieved  enough to protest outside a shop then in a civilised democracy there's not much we can or should do.

On the other [largely left] side there is the view that such companies are not pulling their weight and should contribute more.

Well, I don't agree wholly with this either.

Many (but not all) of the companies criticised have a UK workforce and end up paying considerable sums in employer's tax (AKA employer's National Insurance contributions) at around 13.8% of salary, plus provide employment (useful, right?) and hence generate even more tax through PAYE paid by their employees.

The trouble with corporation tax in a global economy is that it is unfair to many smaller, local firms; in that they cannot afford the set-up costs of an offshore headquarters to launder their profits through.

Competition theory largely states that governments should encourage entrepreneurship and regeneration to keep the markets competitive; a market which makes it hard for new entrants tends to get lazy, with the incumbents carrying on as before, unchallenged.

But non-global new entrants find it hard to challenge the global giants if they end up paying more tax - until they get big enough to avoid tax.  Get the idea? The market becomes skewed against the new entrants.

So one answer - and the left won't like this - is to get rid of corporation tax altogether.

But this creates a new problem.  Not all companies provide such large returns to the exchequer through employment taxes.  E.g. city fund managers may rake-in millions yet employ a handful of staff, whilst large retailers such as Marks and Spencer, Tesco, etc each employ tens of thousands of workers.

So to me the answer appears obvious.  Companies should be allowed to offset their corporation tax bill against their total employer's Class 1 National Insurance contribution.

Essentially many companies making modest profits yet already paying millions of pounds a year through employment taxes would be exempt from paying any corporation tax, yet companies who didn't employ many UK staff would be left with a largely unchanged corporation tax bill.

This would perhaps have a secondary advantage of making it more attractive to employ UK staff, as the employer's tax - widely seen as a disincentive to employment - would be offset against corporation tax.

And, importantly, it would allow smaller, growing UK firms who choose to have UK-based staff to pay corporation tax on a similar rate to the global giants.


Tuesday, 14 May 2013

The Data Plutocrats and a need for a Data Democracy

Yesterday certainly wasn't the first time someone opined the term privacy was counter-productive in relation to data.

"Privacy" is a one-sided open-ended discussion about risk with no consideration of reward.

"Privacy" is an amorphous concept easily spun by proponents of one side or the other.

Privacy: is wholesome, positive, for victims of crime themselves becoming victims of press intrusion; or, privacy gives terrorists and child abusers the space they need to hide amongst us in society.

Discussion about data privacy and related topics could, perhaps, be more constructive if framed as a discussion about balance of power.

After all, privacy primarily concerns us because of our fear that our secrets can be used against us, creating an "information asymmetry" (ht @OrwellUpgraded) that could be abused by the nefarious and amoral.

So should privacy advocates instead be arguing for a data democracy?

Taking a step back, democracy is not a goal in itself.  The end game is a comfortably stable, affluent and sustainable society; which, if one trusts in the inherent good in human nature will itself be a fair and just society.

Similarly, in data terms, we want a society where we are all "data wealthy" - ie have access to information, communications, entertainment; and benefit from the resulting advances in science, medicine, etc only possible through smart use of data.

We want relative stability - a society nimble enough to keep pace with advancing technology, yet resilient enough not to be cajoled into dangerous change.

We want a just and fair society where individuals, corporations and governments can't use our personal data, our everyday secrets, to exert undue control on anyone.

Democracy is probably the best place to start - at least in analysing and attempting to understand the problem.

Today we probably have a data plutocracy, where data power is concentrated in the hands of a few global corporations.

No-one knows for sure whether this itself is inherently dangerous.

Data power has certainly been used for good: the rapid emergence of useful services, the construction of data infrastructure on a truly massive scale, a level of free "social" services.

Data plutocrats like Google provide services like Blogger, which in turn strengthens the power of the individual to challenge traditional autocracies and, for the time being at least, discuss the issues associated with a data plutocracy.

But clearly such concentrations of data power could easily be abused; either by sticky-fingered employees dipping their hands in the data till, by governments, or by corporations themselves in search of profit.

So maybe we should be looking to promote data control structures and data economies that are inherently more democratic.

But how can we go about understanding the data power balance?

I believe we'll find, over time, that many democratic (and economic) concepts are applicable to data.

Already I see a clear left-right political spectrum, at one end "the state" or other controlling force being responsible for administering and apportioning "data fairness" if you like.  The "clean internet" brigade - a worthy cause... But, as we all know, some data animals are more equal than others.  Who governs the governors, who watches the watchmen?

And at the other end, the right-libertarians, who argue the state should not interfere, leaving the question of who will protect the "data weak"?  Who will guard the technologically incapable from losing out when real-world services increasingly rely on the internet?  Who will provide their broadband, guard their personal data, and defend their computers from hackers?

Over the last two decades the data privacy debate has entered the mainstream - that itself is a good thing, but it's now time to move on to talk about the wider issue: a data democracy.


Monday, 13 May 2013

Marks and Spencer latest victim of automated content filtering

Your can forget using Marks and Spencer's free ecard service if your friend is called Dick (or, presumably, lives in Scunthorpe).

After reports that a pensioner was unable to send a card to her friend Dick due to the firm's automated profanity filter, a spokesperson claimed the filter was there "to protect people from harassment" and consequently would not be altering their policy.

Which got me wondering how much protection their automated system offered...

Here's an ecard I sent to myself that didn't get blocked (forgive my childish use of language):

There's a serious point: time and time again I see a reliance on automated content blocking to offer some level of "protection" that impacts legitimate uses of the system whilst being relatively easy to workaround.

In this case I could harass someone simply by spacing-out the insults.

If the service provider then tried to catch insults with spaces in between they would undoubtedly end up blocking even more legitimate uses, like a card to my good friend* Alf Ucker.

I have no issue with companies like Marks and Spencer using such content filtering to warn people against a thoughtless use of offensive language, but to claim such systems are there to protect people from harassment is a stretch - especially since, as I have shown, they can't stop all offensive messages getting through.

(* = imaginary!)


Tuesday, 7 May 2013

The Tech Chasm Series: 1. The Delivery Window

Over the last 20 years technology has changed so many aspects of life it's now hard to imagine a time when flights had to be booked in person at a travel agent and a vandalised phone box was a good enough reason for not letting your parents/partner/butler know you were going to be a bit late home.

But in some areas a technological overhaul is well overdue, leading to a widening chasm between the theoretical capabilities and the practical application of technology.

1. The Delivery Window

The list of items that can't be [legally] ordered online for home delivery can be written on the back of a Royal Mail "I'm sorry we missed you" delivery card, yet few delivery firms provide a delivery window smaller than ten hours.

In fact you're lucky to get word of a delivery before you've missed it, with many firms relying on drivers themselves to give notice of delivery by buzzing your doorbell, before deploying a missed delivery card worded to make you feel like a naughty schoolchild late with a homework assignment:  'You have one more chance to receive these goods'...

I'm sorry for putting your driver out by not waiting by my door for the entire estimated delivery window of 1st May-7th May.   Next time I will help your driver by driving myself to PC World, running the gauntlet of credit offers at 26.3% APR and over-priced extended warranties costing nearly as much as a replacement laptop, only to find what I want is out of stock - but can be dispatched immediately for home delivery...

Now where was I...

Oh yes, once upon a time the avid mail-orderer needed to know the location of just 2 buildings: the nearest "main" Post Office and the local Parcel Force warehouse.

Today, each expedition is a veritable adventure; a 60-mile round trip to the "local" depot for Generic Delivery Services Ltd whose postcode, at least according to any satnav on the market, is just far enough away from the actual building to render your chance of getting to the collection office after work before it closes about as good as getting the full delivery charge refunded from the vendor if the delivery ends up boomeranging back because you failed to collect the package within a week.

That's assuming your package made it back to the depot, unlike the soggy books inside the soggy cardboard box left on my doorstep last October...

Or the pillow thrown over the gate - of the wrong house! Eventually finding its way to its intended recipient a week after the replacement had been delivered - and returned.

Are we to believe that tracking a parcel within these delivery operations is computationally impossible?  At least given the computing power of a ZX Spectrum (with 48k Ram Pack and ZX Microdrive)..

Granted, we have an extremely complex problem: a parcel at location A needs to get to location B, and the recipient R needs to be notified somehow.

You'd have to request their email address and everything...

Plus you'd need to know how long it takes for a parcel to reach the trunk network, travel the trunk network; I mean, it would take some effort to plan shipping routes and schedules (rather than leave them to the driver's discretion, as I assume they  must do now), collect actual data - maybe even build a database if it doesn't all fit on a Microsoft Excel spreadsheet...

And provision delivery vehicles with the latest state-of-the-art satellite technology - or at least ask delivery drivers to keep their phones switched on whilst working.

Maybe NASA could help? After all they seemed to predict the landing time of Curiosity on Mars to a better accuracy than Rural Link Express can track a parcel from Manchester to Woking.

Is it really that hard for a company the size of a small country to divert some of the money it saves avoiding tax in to technology that lets its customers know with reasonably accuracy:

  • What day, at the time of ordering, a parcel will arrive; and, 
  • At some point the day before the time, to within a couple of hours, the parcel will be delivered?
And, instead of marketing this service as an optional extra, provide it to all customers, as it will surely save everyone time and money if a higher proportion of parcels are delivered on first attempt.