Category : Political Economy

Why There’s No Future in Cloud Futures

An inter­est­ing debate took place at last week’s Cloud2020 gath­er­ing regard­ing the via­bil­ity of futures mar­kets for cloud com­put­ing capac­ity. I’m firmly at the skep­ti­cal end of the spec­trum as the title of this post will attest. How­ever, I had not given enough thought as to the rea­sons for my skep­ti­cism. Hav­ing reflected on it a lit­tle I’m more con­vinced than ever that any attempt to cre­ate either pri­mary or sec­ondary mar­kets in cloud com­modi­ties is doomed. In short the lack of fric­tion and a lack of volatil­ity in match­ing cloud com­put­ing sup­ply and demand means there is likely no win­dow for 3rd party mar­ket mak­ers to insert them­selves into this value net­work. Con­tinue Read­ing

The Jobs Are Not Coming Back

The NY Times pub­lished an inter­est­ing arti­cle today the impact of robot­ics on man­u­fac­tur­ing jobs using Amazon’s pur­chase of robot maker Kiva Sys­tems as an exem­plar. The core ques­tion raised in the arti­cle is whether the adop­tion of automa­tion tech­nolo­gies will cre­ate more unem­ploy­ment or whether — as has been seen in the past — dis­placed work­ers will re-train and become employed in other pro­duc­tive areas of the econ­omy. Con­tinue Read­ing

Policy Challenges for A Globally Integrated Innovation, Production and Market Platform

Cloud com­put­ing” is much more than sim­ply a new set of tech­nolo­gies and busi­ness mod­els.  It is rapidly emerg­ing as the plat­form that will under­pin the next gen­er­a­tion of dig­i­tal prod­ucts and ser­vices. Cloud Com­put­ing is trans­form­ing how con­sumers, com­pa­nies, and gov­ern­ments store infor­ma­tion, how they process and exchange that infor­ma­tion, and how they uti­lize com­put­ing power.  Con­se­quently, it opens a new set of pol­icy dis­cus­sions while at the same time under­lin­ing the impor­tance of old debates.

Con­tinue Read­ing

Five Things Politicians Need to Understand About Data Centers

The Office of Gov­er­nor Bob McDon­nell of Vir­ginia issued a press release today trum­pet­ting their suc­cess in con­vinc­ing Microsoft to build its lat­est Gen4 Data Cen­ter in the state. Given huge gulf between the real long-term local eco­nomic ben­e­fit of a data cen­ter like this and the con­tent of the press release I am left to con­clude that either A) The Gov­er­nor really does not under­stand the eco­nom­ics of these things or B) He really does under­stand the true eco­nomic impact and the press release is a cyn­i­cal attempt to fool the voters.

For other politi­cians who might be tempted to spend tax pay­ers money to secure the loca­tions of a data cen­ter here are some things you really do need to understand:

  1. Today’s mod­ern global data cen­ters do not cre­ate any mean­ing­ful num­ber of jobs. Mod­ern global data cen­ters are remotely oper­ated which means that even the few jobs that are cre­ated are not high skilled soft­ware or com­puter engi­neer­ing roles. The type of jobs cre­ated tend to be in the build­ing main­te­nance, HVAC, elec­tri­cal engi­neer­ing and secu­rity domains.
  2. There’s a high price to be paid for secur­ing rel­a­tively few jobs. The fifty num­ber touted by Vir­ginia sounds about right. How­ever, from the Governor’s press release, it sounds like at least $6.9 Mil­lion of direct incen­tives and prob­a­bly an unspec­i­fied amount of long-term com­mer­cial prop­erty tax exemp­tions were used to secure those 50 jobs. Is that a good rate of return?
  3. When a tech­nol­ogy com­pany says they will invest $499 Mil­lion in build­ing a new data cen­ter that does not mean they are invest­ing $499 Mil­lion in your state. Far from it. Most of that cost will be in the com­put­ing infra­struc­ture, com­mu­ni­ca­tions, and other elec­tri­cal machin­ery. That equip­ment is by and large not built in the state where the data cen­ter will be located. In fact a lot of it will be sourced from out­side the coun­try cre­at­ing jobs and employ­ment else­where, but not in your state.
  4. Mod­ern data cen­ters are ‘Shells’ of a build­ing. Microsoft’s Gen 4 data cen­ters employ pre-assembled con­tainer ‘Lego Blocks’ which are rolled onto a con­crete slab and plugged into power, cool­ing and data pipes. If they could get away with­out build­ing any­thing other than the con­crete slab and HVAC con­tain­ment build­ings they would and in years to come they will. This all means that very few con­struc­tion jobs will be cre­ated and cer­tainly not any that require sig­nif­i­cant new up-skilling of your labor force.
  5. On a pos­i­tive note a mod­ern global data cen­ter is one of the most dense power con­sump­tion facil­i­ties around. That is why many sites are selected based on access to hydro-power where the cost of energy and dis­tri­b­u­tion chal­lenges can be kept to a min­i­mum. It is pos­si­ble that the loca­tion of a major global data cen­ter in Vir­ginia will cre­ate the down­stream need to upgrade the state’s elec­tric­ity dis­tri­b­u­tion and gen­er­a­tion infra­struc­ture. Being forced to do this obvi­ously cre­ates ancil­lary ben­e­fits for other company’s but again at what price?

Politi­cians really do need to under­stand the real­ity of these things. It is far too easy to give away the tax pay­ers money today on the assump­tion of major down­stream eco­nomic ben­e­fits tomor­row. In the case of data cen­ters those ben­e­fits are likely to be few and far between.

Why Regulators Need to Mandate Open Data Protocols in the Cloud

TechCrunch’s arti­cle about Intuit’s deci­sion to shut down Quicken OnLine and not let users migrate their data to the new ser­vice Mint.com is an out­stand­ing exam­ple of the emerg­ing prob­lem of ‘Data Lock-In’ and an issue which reg­u­la­tors need to get ahead of before its too late.

As the mar­ket for ‘Cloud’ ser­vices takes off we’re going to wit­ness an explo­sion of new ser­vices whose value propo­si­tion is based on man­ag­ing some aspects of the con­sumers life on-line. The big buck­ets already are life-centric ser­vices such as health and fit­ness, per­sonal finances and of course man­ag­ing you social graph. The expand­ing ros­ter of new ser­vices will be accom­pa­nied by an equally large num­ber of fail­ures. That is the nat­ural order of things in new mar­kets. Not all new entrants will find a busi­ness model which works or is sus­tain­able over the long term. Reg­u­la­tors need to start think­ing about the con­sumer pro­tec­tion regimes that need to be put in place as this happens.

I was struck by some of the debate sur­round­ing the Intuit issue by the num­ber of folks who think its not a big deal because Quicken On-Line is a ‘Free’ ser­vice. That could not be fur­ther off the mark. One of the chal­lenges for any new entrant in the crowded mar­ket for ‘Cloud’ ser­vices will be fig­ur­ing out how to make their ser­vice as ‘Sticky’ as pos­si­ble. After hav­ing invested a great deal to acquire new cus­tomers you want to make sure its not easy for them to leave or switch to a com­peti­tors ser­vice (Ever won­der why its incred­i­bly dif­fi­cult to dump you mobile provider? It’s called ‘Churn’ and ‘ARPU’)

One way to build ‘Stick­i­ness’ is to ensure that new cus­tomers have to invest a sig­nif­i­cant amount of per­sonal effort to extract the full ben­e­fits of the ser­vice. In Quicken Online’s case that was the time required to upload bank­ing infor­ma­tion and months and months of enter­ing indi­vid­ual finan­cial trans­ac­tions. Mak­ing it easy for cus­tomers to export their data from the ser­vice would under­mine ‘Stick­i­ness’ and there­fore increase the likely hood that cus­tomers might switch to a competitor’s ser­vices. In the highly com­pet­i­tive ‘Cloud’ ser­vice mar­ket it will be very tempt­ing to find ways to bind users to ser­vices by mak­ing it dif­fi­cult for them to take their data and run.

Con­sumers need to be edu­cated about the dan­gers of ‘Data Lock-In’ so that they increas­ingly demand Open Data Access (ODA) from ‘Cloud’ ser­vice providers. I think its also about time that gov­ern­ment reg­u­la­tors started think­ing about man­dat­ing ODA for all ser­vice providers. These reg­u­la­tions would be a nat­ural exten­sion of exist­ing pri­vacy reg­u­la­tions by enshrin­ing the prin­ci­ple that per­sonal infor­ma­tion of any sort belongs to the con­sumer, not the provider, and the con­sumer has the right to take that data with them when they leave a ser­vice. This would be a direct ana­log to ‘Num­ber Porta­bil­ity’ require­ments in the mobile tele­phony sec­tor. Ser­vice providers would be required to ensure that all con­sumer owned data can be exported from their ser­vice using open stan­dard for­mats and protocols.

Such a reg­u­la­tory require­ment would pro­tect con­sumers of the sort of mess now faced by Inuit’s cus­tomers. It would also force ser­vice providers to focus on broader value propo­si­tion of their ser­vice rather than rely­ing on ‘Data Lock-In’ as a com­pet­i­tive lever. This focus on ‘Func­tional’ com­pe­ti­tion would be a great thing for con­sumers and the market.

Does Economic Growth Require a Growing Population?

World Population 1800-2100

World Pop­u­la­tion 1800–2100 show­ing alter­na­tive growth projections

The UK’s Royal Soci­ety has just announced a major new research project on the impli­ca­tions of pop­u­la­tion growth. The study is to be headed by the Nobel Lau­re­ate, Sir John Sul­ston. As the RI points out pop­u­la­tions stud­ies is quite a cycli­cal area of research. The ‘60s and ‘70s were a boom time for the field with less focus being given the the sub­ject in the ‘80s and ‘90s. The fact that the topic is mov­ing back up the polit­i­cal agenda prob­a­bly has much to do with the inter­sec­tion of pop­u­la­tion on two key pol­icy issues; eco­nomic growth and cli­mate change.

The polit­i­cal debate about pop­u­la­tion growth has being going on for nearly three hun­dred years. Thomas Malthus, seen as the father of demog­ra­phy, clearly held the view that expo­nen­tial pop­u­la­tion growth would out­strip improve­ments in agri­cul­tural pro­duc­tion, con­demn­ing economies to a per­ma­nent state of sub­sis­tence. Adam Smith on the other hand believed that eco­nomic growth and improve­ments in people’s socioe­co­nomic cir­cum­stances would reduce fer­til­ity rates over time.

World Population Growth Rate

World Pop­u­la­tion Growth Rate Over Time

There’s a prima facie case sup­port­ing Smith’s asser­tions. The fer­til­ity rate of many of the world’s most devel­oped economies (Much of Europe, UK, Japan etc. but notably not the United States due in large part to the impacts of immi­gra­tion) has tum­bled over the last forty years. That decline has mir­rored a period of mas­sive eco­nomic growth and ris­ing liv­ing stan­dards in these coun­tries. The ‘Baby boom’ has become the ‘Baby bust’ with very sig­nif­i­cant impli­ca­tions for affected geo­gra­phies. As the demo­graphic mix inverts there are fewer and fewer younger work­ers able to carry the social cost of the ris­ing num­ber of elderly in the population.

Inter­est­ingly, even though the total global pop­u­la­tion is still grow­ing the rate of growth has in decline since the mid-sixties. There is a diver­sity of research which shows that fer­til­ity rate has an inverse rela­tion­ship with fac­tors like edu­ca­tional level (Par­tic­u­larly in Women) and income level. If the RI study is to be use­ful in guid­ing pol­icy mak­ers then it needs to deliver some defin­i­tive research on which fac­tors really do impact fer­til­ity rates, and in what manner.

There is a per­spec­tive (One I do not sub­scribe to) that a grow­ing global pop­u­la­tion is required to sus­tain the needed lev­els of eco­nomic growth. The argu­ment is put for­ward that con­tin­ued growth in the pro­duc­tion of man­u­fac­tured goods will require and expand­ing work­force to make those goods. This gets to the heart of one of my per­sonal inter­ests regard­ing the impact of tech­nol­ogy on eco­nomic growth. I would argue that the oppo­site sce­nario is more likely i.e. that even an increas­ing sup­ply of man­u­fac­tured goods will requires less, not more work­ers over time.

The num­ber of work­ers now employed in build­ing cars in the most advanced and auto­mated fac­to­ries today is dimin­ish­ingly small com­pared to just thirty years ago. With con­tin­ued advances in robot­ics, process con­trol tech­nol­ogy and infor­ma­tion sys­tems this trend is likely to con­tinue. Fifty years from now its likely that many fac­to­ries will ingest raw mate­ri­als at one end and spit out fin­ished good at the other with very few human work­ers required in-between.

To date this trend has been coun­tered by the rel­a­tive ease with which man­u­fac­tur­ing facil­i­ties can be trans-located into low wage economies. How­ever, the very act of intro­duc­ing man­u­fac­tur­ing into low wage economies has the effect of rais­ing liv­ing stan­dards and worker expec­ta­tions which in turn leads to demands to higher wages. When com­bined with worker short­ages this can place a sig­nif­i­cant upward pres­sure on wages under­min­ing the location’s orig­i­nal cost advan­tage. China is now wit­ness­ing this phe­nom­e­non with many man­u­fac­tur­ers now look­ing to move to other lower wage cost East Asian economies. In my view this will over time lead to a lev­el­ing of the global wage cost ‘Play­ing field’. As we reach that point the pres­sure to inno­vate in man­u­fac­tur­ing automa­tion will become extreme because that will be the only way to avoid being caught between the ‘Pin fac­tory’ and the ‘Invis­i­ble hand’. As global man­u­fac­tur­ing com­pe­ti­tion shifts from try­ing to find the low­est wage econ­omy to find­ing the cheap­est way to man­u­fac­ture through automa­tion we will see a dra­matic reduc­tion in the global man­u­fac­tur­ing work force.

The chal­lenges with this sce­nario are daunt­ing. If there is truly a cor­re­la­tion between ris­ing income lev­els and declin­ing fer­til­ity rates then a long period of tech­nol­ogy lead effi­ciency improve­ments in man­u­fac­tur­ing and its con­comi­tant pos­i­tive impact on eco­nomic out­put should cor­re­late with a con­tin­ued slow­ing of the global pop­u­la­tion growth rate. If the point is reached where global pop­u­la­tion growth rate becomes neg­a­tive then you enter into the poten­tially ‘Utopian’ sit­u­a­tion where a smaller and smaller global pop­u­la­tion shares an ever grow­ing eco­nomic pie. Alter­na­tively, if the global man­u­fac­tur­ing work force con­tin­ues to shrink due to automa­tion while the pop­u­la­tion con­tin­ues to grow and those lost man­u­fac­tur­ing jobs can not be more than off­set by growth in ser­vices sec­tor employ­ment then you have a polit­i­cal time-bomb on your hands.

I seri­ously hope the RI study will look at the the cor­re­la­tion of pop­u­la­tion dynam­ics and eco­nomic growth and will fac­tor in the role that tech­nol­ogy will play on both sides of the equa­tion. The likely impact of cli­mate change on these issues I will leave for another post.

The Reality of ‘Private Cloud’

Mary-Jo Foley posted a provoca­tive arti­cle yes­ter­day about the real­ity of ‘Pri­vate Cloud’ offer­ings and who is dri­ving the demand; cus­tomers or vendors?

As I was one of the peo­ple respon­si­ble for design­ing Microsoft’s Pub­lic Sec­tor ‘Cloud’ strat­egy I have some opin­ions about this issue which I wanted to share.

It’s fair to say that until we started look­ing at the world­wide cus­tomer require­ments for ‘Cloud’ in the pub­lic sec­tor the whole issue of ‘Pri­vate Clouds’ was not a major part of the company’s over­all ‘Cloud’ strat­egy. How­ever, once you start look­ing at the require­ments of pub­lic sec­tor orga­ni­za­tions out­side the US you very quickly real­ize that stan­dard ‘Pub­lic Cloud’ offer­ings will not cut it.

The major­ity of for­eign gov­ern­ments have data sov­er­eignty reg­u­la­tions which pro­hibit the stor­age and trans­port of data beyond the country’s bor­ders. Many gov­ern­ments also have very seri­ous con­cerns about the reach and impli­ca­tions of the US Patriot Act which requires any US based ‘Cloud’ provider to dis­close any data held within their sys­tems to the US Gov­ern­ment, upon request, no mat­ter where that data is stored. Obvi­ously these require­ments do not affect the pro­vi­sion of ‘Cloud’ ser­vices to the US gov­ern­ment. How­ever, providers will still need to ensure that US gov­ern­ment data is not ‘Smeared’ across the provider’s global ‘Cloud’ infra­struc­ture and is instead kept within data cen­ters hosted in the US.

On face value these require­ments might be seen as a com­plete bar­rier to the adop­tion of ‘Cloud’ by for­eign gov­ern­ments. How­ever, these chal­lenges are bal­anced by a huge pres­sure to improve the effi­ciency of IT pro­vi­sion across the Pub­lic Sec­tor. The extreme bud­getary pres­sure being faced by many gov­ern­ments is forc­ing a re-evaluations of how IT ser­vices are deliv­ered and at what costs. In this light the costs advan­tages of ‘Cloud’; scale, elas­tic and auto­mated pro­vi­sion­ing, pay-as-you-go, reduc­tion in cap­i­tal expen­di­ture and con­sol­i­da­tion of oper­a­tions etc. are all highly attractive.

The only way to ‘Square the Cir­cle’ is to offer a ‘Pri­vate Cloud’ solu­tion i.e. a set of tech­nolo­gies which will let gov­ern­ments imple­ment IT infra­struc­ture which has ‘Cloud’ attrib­utes but which can be kept sep­a­rate from the ‘Pub­lic Cloud’ infra­struc­ture and com­pli­ant with the country’s required pol­icy, reg­u­la­tory and secu­rity regimes.

The UK is a good exam­ple. The UK gov­ern­ment is expect­ing all the advan­tages of ‘Cloud’ with­out the expo­sure of putting UK gov­ern­ment data into the ‘Pub­lic Cloud’ infra­struc­ture with all the expo­sure that implies. At the end of the day ‘Cloud’ is a par­tic­u­lar approach to sys­tems and work­load ‘Man­age­ment’ that deliv­ers the ben­e­fits I’ve out­lined above. Whether these ben­e­fits are deliv­ered within a pri­vate dat­a­cen­ter or across a pub­lic infra­struc­ture is really immaterial.

From a com­pet­i­tive per­spec­tive its also impor­tant to under­stand why play­ers like Ama­zon and Google want to play down the rel­e­vance of ‘Pri­vate Cloud’. The pub­lic sec­tor is a very large and impor­tant IT mar­ket. In many coun­tries the gov­ern­ment is the sin­gle largest spender on IT. Ven­dors who only offer ‘Pub­lic Cloud’ ser­vices are fail­ing to meet the most basic needs of pub­lic sec­tor cus­tomers out­side the US. If you want to play in the pub­lic sec­tor mar­ket around the world you will need both a ‘Pub­lic’ and  ‘Pri­vate Cloud’ strategy.

It was inter­est­ing to see that hav­ing dri­ven this set of require­ments out of the pub­lic sec­tor side of the busi­ness the con­cept of ‘Pri­vate Cloud’ started to find sig­nif­i­cant trac­tion in the enter­prise seg­ment. There are plenty of large pri­vate sec­tor com­pa­nies who are not yet ready to move their sen­si­tive data into the ‘Pub­lic Cloud’ and yet want the ben­e­fits of ‘Cloud’ work­load man­age­ment to drive effi­ciency in their IT ser­vice pro­vi­sion. In my view most large enter­prise orga­ni­za­tions will end up using a blend of both ‘Pub­lic’ and ‘Pri­vate’ cloud.

IT Platforms and the Ecology of Innovation">IT Platforms and the Ecology of Innovation

It has been fun being back in Bei­jing this week. My first visit to the city was in 1989 when the streets were still filled with bicy­cles. I’ve been back here about once a year since then yet the pace of progress on so many lev­els is still hard to com­pre­hend. Growth of the ‘Star­bucks Index’ alone is mind blow­ing and seems to be reach­ing Seat­tle or New York den­si­ties. The other thing I’ve noticed is the incred­i­ble focus on effi­ciency; from the three hour turn-around for a visa at the Chi­nese Con­sulate in Zurich (Try that at the US Embassy), to the immi­gra­tion process and flow through the air­port. Yes, the traf­fic is still a night­mare and seems to be get­ting worse. I don’t live here so I can imag­ine there is an alter­nate real­ity I’m not aware of. However, the gen­eral expe­ri­ence for vis­i­tors is pretty impressive.

I had a chance to catch up with Microsoft’s National Tech­nol­ogy Offi­cer Sean Zhang on Wednes­day for a great evening of con­ver­sa­tion. I’ll also be able to see Peter Moore and Michael Thatcher before I leave so that’s an added bonus.

The rea­son for the visit to Bei­jing was an invi­ta­tion to present at the Chi­nese Acad­emy of Gov­er­nance today. The Acad­emy is respon­si­ble for train­ing senior civil ser­vants in the PRC admin­is­tra­tion. My pre­sen­ta­tion is titled “IT Plat­forms and the Ecol­ogy of Inno­va­tion” (PDF copy here) and focuses on the how the evo­lu­tion of IT plat­forms has enabled the devel­op­ment of the global ser­vice econ­omy. I’ll take a his­toric look at how these plat­forms have evolved and then will talk about the com­ing dis­rup­tive effects of the ‘Cloud’ plat­form and those which follow.

This is pre­sen­ta­tion comes out of ongo­ing work with Prof. John Zys­man and the team at UC Berkeley’s BRIE. One of the things which is cen­tral but still the cen­ter of a heated debate is how you define an IT Plat­form i.e. rigorously enough for it to be used as an ana­lyt­i­cal def­i­n­i­tion when look­ing at the sur­round­ing polit­i­cal econ­omy issues. The cur­rent def­i­n­i­tion I’m using is:

A con­sis­tent devel­op­ment envi­ron­ment sup­ported by new soft­ware and hard­ware archi­tec­tures, based on stan­dards and avail­able at scale, that enables ser­vice and busi­ness model innovation

This is not per­fect and I’m open to sug­ges­tions about how to improve it. Any def­i­n­i­tion needs to be able to delin­eate his­toric plat­form tran­si­tions in a clear and defen­si­ble way and also be able to help iden­tify when a new tran­si­tion is tak­ing place.

I’m look­ing for­ward to a inter­est­ing dis­cus­sion and debate about the oppor­tu­ni­ties and pol­icy chal­lenges these new plat­forms will cre­ate in the Chi­nese con­text. Should be fun.