Embed Size (px)
DESCRIPTION
Guest lecture on open government data, for students of Saxion University for Applied Sciences
Citation preview
Open (Gov) DataThe What, Why, and How in 50 examples or less
Guest lecture, 30 October 2014
Ton Zijlstra, @ton_zylstra
Lessons from…
Enschede
Amsterdam
European Commission
World Bank
Province North-Holland
Leeuwarden
emerging patterns
Province Utrecht
Schiedam
European Space Agency
High Court of Audit
Min Interior
Min Economics
Dutch base registers
Danish base registers
Moldovan gov
Heerlen
Flemish Chancellary
Min Housing (DK)
Min Fin (DK)
ePSIplatform
LAPSI
hansjebrinker.com
impacts infrastructure, mobility, coastal defence, costs
http://hansjebrinker.com
&What it is
How to do
Why do it &
What it is
Data that is
Gathered for a public task !
Pro-actively published !
To be used by others
No legal barriers
Everyone has equal access !
Open licensed (PD, or attribution) !
No statement of interest !
No usage restrictions
No tech barriers
Open standards !
Findable !
In bulk and/or in pieces !
Raw and timely
No monetary barriers
Free of charge !
Marginal costs at most
untapped abundance
https://secure.flickr.com/photos/opensourceway/5535034664
towards by default, by design
Why do it
Market and civic domain !!!!!!!!!!!!!!!
Government domain !!!!!!!!!!!!!!!
There always was value in data
Government Data
Re-user 1
Re-user 2
Open Data
2: effect of ‘Open’ !- free or not? - - ∆ demand - Ev price !
- ∆ fiscal revenue > ∆ sales revenue ?
!OXFORD 2009 POPSIS 2011
1: Market value PSI !!!!!!!!!
PIRA 2000 (€95 B) MEPSIR 2006 (€ 37 B)
3: cost of ‘Open’ <<1% of costs
Market and civic domain !!!!!!!!!!!!!!!
Government domain !!!!!!!!!!!!!!!
Gov body 1 DataOpen Data
Government - Interaction w market/citizens changes
- Quality of data improves - Increased efficiency & effectiveness - public tasks shift (smaller, different)
Digitization has shifted that valueMarket
- Barriers to entry drop away
- market dynamics: chains change
- paradigm shift: from owning to using
Value looks different: - less lineair - not monetary - more equally spread - hard to measure - hard to correlate - but doable
four valuable reasons
more efficient government !
better public services !
more transparent government !
new socio-economic value
four valuable reasons
more efficient government !
better public services !
more transparent government !
new socio-economic value!
http://www.flickr.com/photos/59937401@N07/5858059202/
DK addresses 2010: value > 70 * cost !
ETLA, SME’s geo data grow 15% faster !
Spain 2011/12, up to 600 million Euro !
POPSIS 2012, 21 cases !
Vickery / EC 2012, 2% GDP EU !
McKinsey 2014, $3 trillion+ !
„Open Data for Economic Growth” (WB, june 2014)
!!
all empirical evidence points same way
open data in the EU
Public, unless!
openbaar, tenzij!
openbaar, tenzij!
openbaar, tenzij!
openbaar, tenzij!
openbaar, tenzij!
openbaar, tenzij!
35 yrs FOIA, 10 yrs PSI Directive, 6 yrs open standards
four valuable reasons
more efficient government !
better public services !
more transparent government!!
new socio-economic value
various reasons, a transparant sector results
284 organisations, aidtransparency.net
http://www.flickr.com/photos/59937401@N07/5858059202/
economic value? transparency?
not my departments job!
four valuable reasons
more efficient government!!
better public services !
more transparent government !
new socio-economic value
29
participatory budgeting
30
government is a strong re-user: 6.5m GBP saved
Denmark projects 26M+ Euro annual savings
BC #1 user, 33% of downloads
open data shows 200m GBP potential savings
four valuable reasons
more efficient government !
better public services!!
more transparent government !
new socio-economic value
35
http://wiki.openstreetmap.org/wiki/File:Haiti_earthquake_damage_map.png
http://hot.openstreetmap.org/
allowing disaster response
http://www.theguardian.com/society/2013/mar/12/nhs-transparency-open-data-initiative
1000 less heart surgery deaths / year
https://secure.flickr.com/photos/76652722@N04/6878044757
pupil and elderly public transport: better = cheaper
Uganda: Open Data and Community Health Monitoring
28
z 33% reduction in under-5 mortality
z 20% extra utilisation of out-patient services
z Significant improvements in: z Immunization z Waiting times z Absenteeism
source: Andrew Stott, World Bank
policy problem
new ways
four valuable reasons
more efficient government!!
better public services!!
more transparent government!!
new socio-economic value!
Not just gov
42
44
encourage what you cannot do yourself
How to do
Costs too much; What’s the business case; Has commercial value; Possible privacy issues; Confidential info; It’s not ours, and we don’t know who owns it; It’s not ours and supplier won’t allow it; The quality isn’t very good; We don’t know where it is; Not our job; It’s in a useless format anyway; I don’t have the authority; People will misuse the data; People will use it wrongly; Only we understand our data; We’ll get sued; Files are just too big; Too little bandwith; It starts with this, but where’s the end? It’s there, but can’t be opened; Data is dated/too old; It’s not in digital format; Is this even legal?; Our Minister says no; We never have done this before, why start now?; I don’t see the use; Nobody will be interested; No time; No resources; Just do FOIA requests; We’ll publish it redacted; It’s not complete; It contains errors; It’s commercially sensitive; Combining this with other data is dangerous; People will come to wrong conclusions; People will get lost and confused; It will trigger endless discussions; We can’t confirm or deny we have that data; We’ll get feedback, and can’t handle that; Our IT supplier says it’s not possible; Our IT supplier will charge too much; Our site will crash; It’s already online! (but in unfindable PDFs); If people download it and use it later it will be outdated; I can’t take responsibility for all the reuse; People will get angry; Our data is in contradiction to the data of the department that is in charge of the topic; Only we truly understand.....statistics/meteo/geo/laws; We’ll disrupt the market; It will only be used to attack us.
many shades of ‘no’
Слишком дорого; В чем выгода; Есть коммерческая ценность; Возможные проблемы с личными данными; Конфиденциальные данные; Эти данные не наши, мы не знаем, кому они принадлежат; Мы не владеем данными, а поставщик не разрешит; Качество не очень хорошее; Мы не знаем, где они; Не наша работа; Они в бесполезном формате; Я за это не отвечаю; Люди воспользуются данными неправильно; Люди используют данные в неправильных целях; Только мы понимаем наши данные; Нас засудят; Файлы очень тяжелые; Низкая скорость загрузки; Данные устарели; Данные не в цифровом формате; Это вообще законно?; Министр запретил; Мы никогда этим не занимались, зачем начинать?; Я не вижу в этом пользы; Никому не будет интересно; Нет времени; Нет ресурсов; Мы будем отвечать только на запросы, связанные со свободой информации; Мы опубликуем информацию выборочно; Данные неполные; В файле есть ошибки; Это коммерчески важная информация; Если объединить эти данные с другими, они могут быть опасными; Люди сделают неверные выводы; Люди запутаются; Начнутся бесконечные обсуждения; Мы не можем ни подтвердить, ни опровергнуть, что у нас есть эти данные; Начнется обратная связь, у нас нет возможности всем отвечать; поставщик наших ИТ систем говорит, что это невозможно; Наш сайт упадет; Эти данные уже онлайн! (но в PDF, и их невозможно найти); Если люди загрузят их сейчас, а используют потом, они устареют; Я не могу отвечать за всех повторных пользователей; Люди разозлятся; Наши данные не совпадают с данными ведомства, которое отвечает за этот вопрос; Только мы по-настоящему понимаем … статистику, погоду, законы; Мы нарушим баланс на рынке; Ими воспользуются, чтобы нас атаковать.
Множество оттенков «нет»
census.okfn.org
Mins I&E / EA 2015, e.g. NDW 600M records / day
100% CC0, 90% dode links
April 2014 Oct 2014
http://epsiplatform.eu/content/european-psi-scoreboard
OGP Action plan
ready, but for what?
real barriers: it’s a transitionSee http://www.flickr.com/photos/epsiplatform/5737203950/
55
top
bottom
middle
56
top down leadership
Networked life, networked work, networked learning
know the gov does not exist. find 1 civil servant
bottom up trailblazing
Actief uitnodigend
community engagement
http://www.flickr.com/photos/dhammza/492882480/
iinformal is important
61
middle
middle out organizing
where to start?
1 Core reference data
accessibility
65
2 Data people care about
cleaning up the country side
3 Data gov cares about
OPEN data
open DATAvs
postcodes….if you darehttp://www.zylstra.org/blog/2013/10/theoretically-open-post-codes-not-so-much-in-practice/
2 years ago
now
DK Roadmap: open is key ingredient
serious issues
serious data=
health care? gov spending?
„just add openness”
Now
78
instru ment
inter vention&
value, efficiency, impact
use open data as a policy instrument
policy issue
stakeholders open data
!• Texel • Hollands Kroon • Heerhugowaard • Purmerend • Velsen • Haarlem • Medemblik • Den Helder • Schagen
!• tourism • real estate • neighborhoods • public spaces • flash floods • local entrepreneurs • school/elderly transport • disused shops • tbd
9 local communities & themes
#pnhslim
open data as intervention
flash flood prevention with greener gardens
compare local spending
83
<<1%
rule of thumb
84
how to change existing financing models?
https://secure.flickr.com/photos/alexjbutler/14953247679/
Who is to pay the bill? 5 options
Data setsRegistration+ updating
Delivery to users
Public sector
Private sector + society
Reporting fee
Registration fee
User fee
User fee
State
budget
Who is to pay the bill?
Data setsRegistration+ updating
Delivery to users
Public sector
Private sector + society
Reporting fee
Registration fee
User fee
User fee
State
budget➢Discourage use
➢Affect quality data
➢Discourage use ➢ govt pays govt ➢Admin costs
➢If there is a concrete return ➢If admin costs low
➢Discourage use ➢Miss out on value
Some Dutch figures on income key registers 2011 * 1 million euros
Registration fees
Public sector use
Private sector use
Own re-use activities
State budget Total
Buisness registers
67 6 42 5 - 120
Cadastral registers
130 15 35 30 - 210
Topogra-phical map
- 9,5 0,5 - 14 24
Adresses - - - - 4 4
Total 197 30,5 77,5 35 18 358
embedding into internal policy?
leadership commitment !
procurement procedures !
procurement contracts !
IT specs & change management !
policy plan writing !
project plan writing
https://secure.flickr.com/photos/driek/2938311931
internal policy addresses data holder concerns
open by design
privacy by design&
http://www.flickr.com/photos/mdid/3271972434/
too important to just fix at the end
treating openness as an incident
#pnhslim
mission, infrastructure, operational
Issue driven
Tech drivenvs
96
data pitfalls
What, but not why Evidence based: correlation / causation Direct re-use vs proxies Pre-hypothesis tool: q’s, context, experiments Big data also Σ small data (access / ownership)
97
seeing impact
http://www.flickr.com/photos/chipdatajeffb/8421625784/
yay! impact!
http://www.flickr.com/photos/chipdatajeffb/8421625784/
where is the next Google OR where is the killer app for open data
„where’s the first 1B+ exit?”
making money count
„We’re not an open data company”
costs 500k? or yields 3M5 & 100 jobs?
http://www.flickr.com/photos/chipdatajeffb/8421625784/
yay! impact!
http://www.flickr.com/photos/chipdatajeffb/8421625784/
if you know why, you’ll see impact
http://www.flickr.com/photos/ringroadproductions/144808172/
CreditsOpen Data!
The What, Why, and How in 50 examples or less
Photos: all CC BY NC SA Ton Zijlstra, except where mentioned on photo (CC respective authors) !Slides: CC BY NC SA except where source stated downloads at http://slideshare.net/thegreenland
Ton Zijlstra, @ton_zylstra, [email protected]
