Government agencies are using the power of analytics to understand government performance as well as analyze key trends, catch fraud, and drive better citizen engagement. In this session, you will learn tips on using data to effectively do your job better. Learn key analytical strategies that will help you become an analytical star within your agency or organization.
4. Data.gov
• Open government flag
project for the
administration
• Provides instant access
to over 450,000 datasets
in easy to use formats
• Contributions from UN,
World Bank, and 172
agencies
• Encourage development
of innovative applications
• Drive innovation and
“A Strategy for American knowledge use across
Innovation” published
September 2009
the globe
May 2012 4
5. Be an Analytics Star
The smartest people don‟t work for you
Harness the energy of the public to do amazing things
May 2012 5
6. Creating a Data Ecosystem
1. Make data discoverable
– and give it freely to developers,
scientists, and citizens
2. Connect the community
– to allow collaboration through
social media, events, platforms
3. Provide an infrastructure
– built on standards
– backed by APIs
4. Encourage developers
– to create apps that empower
people‟s choices
– with easy documentation
5. Tell data stories
– of what others have done/can do
May 2012 6
7. Easy to Find Data
• Keyword search
• Faceted search
• Filter by category
• Filter by type
May 2012 7
8. Open Communities
Community
Health ✓
Law ✓
Energy ✓
Education ✓
Ocean ✓
Safety ✓
BusinessUSA ✓
Manufacturing ✓
Open Data ✓
Semantic Web ✓
Developers ✓
Ethics ✓
Cities
Research and Development
Human rights
+ many more…
May 2012 8
9. Cool Things Communities
Highlight
• Data that community experts have deemed best-of-
the-best
• Apps that transform data in understandable ways to
help people make decisions
• Challenges inspire citizens to create new and
unthought-of ways to visualize and present data
• Blogs that tell stories about uses of data
• Maps providing interesting visualization ideas and
ways to reuse those map services
May 2012 9
10. Challenges Spark Ideas
• Energy.Data.gov
connects works
with challenges
across the nation
to integrate
federal data and
bring government
personnel to
code-a-thons
• www.challenge.gov
May 2012 10
11. Data APIs
• API Catalog
• API Key Management
– Register once, use everywhere
across the Federal government
– Basic API Key Auth
• Central Proxy capability
• Key validation service
– Basic application statistics
• Developer Documentation
– Use of templates describing parameters
May 2012 11
12. Open Government Platform
• www.opengovplatform.org
• Bilateral partnership with
Government of India’s
National Informatics Centre
• Open source version of Data.gov
• Offered free (as in speech and beer) to any city or
nation
• Government participation in Open Source Software
July 28, 2011 12
13. Open Government Platform
(cont’d)
• Public commits on Github
• Public bug tracker
• Public mailing list where discussion happens
• Source code on Github
– https://github.com/opengovplatform
• US Data.gov has implemented a component of the
platform already onto its site
• Coordinating with other open data platforms and
data providers
July 28, 2011 13
14. Managing Open Data Tomorrow
• Bring data up and out of government to the public ★
• Make data accessible and linked ★★★★★
• Provide simple ways to visualize the data
• Create communities to understand and apply data
• Connect and collaborate with small businesses, industry,
and academia to drive innovation
• Develop open source open government data platform with
India for global use and further community development
• Share with others to understand global issues
Be the change you want to see in the world
– Ghandi
May 2012 14
15. Be Change Drivers!
1. Get others in your agency excited about making
interesting datasets public
2. Create a challenge of your own to launch innovation
3. Get connected with the Data.gov communities -
there are over 22 now!
May 2012 15
16. Let’s work together to set the
data free!
www.Data.Gov
christopher.musialek@gsa.gov
@usdatagov
@cmoose
May 2012 16
17. LETTING DATA LEAD DECISIONS
Improving Government Performance at the Local, State and
Federal levels
Carter Hewgley
FEMASTAT Director
Department of Homeland Security July 26, 2012
18. LOCAL PERSPECTIVE
Promoting Inclusion at DC Public Schools (DCPS)
In 2009, ~26% of special
education students in DCPS
were being educated in the
most “restrictive environment”1
In FY 2010, the District spent
$154 million on private tuition
for these students2
19. DCPS reduced non-public enrollment by
Non-Public Enrollment
improving and analyzing student data, and declined by 23% or 596
students in two school
then targeting interventions toward trends years through:
and needs3 • Increased graduation rates
among older students
• Preventing younger
At the beginning of SY 2009, placements through more
inclusive practices at their
DCPS was not tracking home schools
enrollment data on non-public
students – so we built a
tracking system and started • Improved accuracy of
studying the data. enrollment records
20. Reduced non-public enrollment created
significant savings, which could be
redirected toward inclusive services within
public schools 3,4
DCPS began comparing enrollment
and expenditure data to reduce Expenditures on Non-
Public Tuition dropped
errors, waste, fraud and abuse by $19 million or 12%
between FY 2010 and FY
2011
21. STATE PERSPECTIVE
Reducing HIV Transmissions in the District of
Columbia
In 1999, the CDC recommended all states use
confidential names-based reporting of HIV
testing/transmission to improve surveillance
accuracy 5
As late as 2006, DC Government had not
transitioned to names-based reporting of HIV
transmissions
Based on AIDS data, it was estimated that 5% of
the District‟s population had HIV or AIDS in 2006 –
the highest infection rate in the nation
22. Until 2007, DC government‟s response DC officials
to the HIV/AIDS Epidemic was not knew AIDS
prevalence was
driven by sound epidemiological higher than any
surveillance on HIV transmissions6 other city or
state, but did
not empirically
HIV Cases, Deaths, and Prevalence in DC know new
2000 16,000 infections were
Newly Diagnosed Cases
12466 14,000 on the rise
Living HIV Cases
11,552
1500 12,000 Newly Diagnosed
10,000
1332 Deaths
1000 8,000
1103
6,000 Living HIV Cases Many assumed
500 4,000 it was a
399 423 2,000 “younger”
0 0 epidemic
2006 2007
600 New HIV Cases in DC by Mode of Transmission Federal law
500
prohibited DC
from
Newly Diagnosed
MSM/IDU
400
IDU implementing
Cases
300
Risk Not Identified
needle
200
Heterosexual Contact
exchange
100 MSM
programs
0
2006 2007
23. As HIV transmission data became more accurate, the District
started evidence-based interventions that slowed
transmission rates considerably6
New HIV Cases in 2010 (by age)
HIV Cases, Deaths, and Prevalence in DC
2000
15,000
Living HIV Cases
3%
New HIV Cases
5%
1500
10,000 15% 13-19
1332
1000 1103 1149 29% 20-29
853 835 5,000 30-39
500
399 423 40-49
343 288 207
0 0 24%
50-59
2006 2007 2008 2009 2010
>60
24%
Newly Diagnosed Deaths Living HIV Cases
New HIV Cases in DC by Mode of Transmission
600 Among new cases, 20-
500
somethings account for
the largest share of
New HIV Cases
400 MSM/IDU
new infections – so
IDU
300
Risk Not Identified
prevention strategies
200 Heterosexual Contact
must account for
100 MSM behavior trends in this
0
age cohort.
2006 2007 2008 2009 2010
24. District youth age 15-19 are at growing risk of transmitting chlamydia and
gonorrhea, so understanding transmission behavior is key to preventing
further HIV prevalence in that age cohort.6
DC Chlamydia Cases by Age and Year
100% 191 311 368 306 280
350 641 693 658 495
80% 465 865 1007 878 712
>40
60% 1985 1980 1650 30-39
1041 1895
25-29
40%
20-24
2694 2610 2351 15-19
20% 1239 2215
0-14
0%
2006 2007 2008 2009 2010
DC Gonorrhea Cases by Age and Year
100% 260 230 167
216 267
80% 324 369 364 328 244
310
>40 9.5% of the District‟s
370 365
60% 299 368 30-39
25-29
population is age 13-
732 600
40% 512 693
712
20-24 19, and interventions that
20% 495 638
880 871 743
15-19
13-14
prevent gonorrhea and
0% 0-12 chlamydia may also
2006 2007 2008 2009 2010 lessen HIV prevalence
25. FEDERAL PERSPECTIVE
Improving Employee Satisfaction at the Federal
Emergency Management Agency (FEMA)
In 2010, FEMA ranked #206 out of 224 agencies
on the Best Places to Work (BPTW)
In 2011, FEMA‟s position fell to #231 out of 240
ranked agencies 7, 8
Are you Considering Leaving in the
DHS & FEMA BPTW Score Next Year?
100%
58.6 56.6
54.4 56.2 55.9 41%
49.1 49.8 51.1
48.3
80%
47.5 YES
60%
40% 71% 67% 59%
20%
No No No
0%
2005 2007 2009 2010 2011 Government DHS FEMA
No
DHS FEMA Yes, to take another job within the Federal Government
Yes, to retire
Yes, other
Yes, to take another job outside the Federal Government
26. FEMA uses a performance management process
called FEMASTAT to evaluate progress against
strategic objectives – improving employee
satisfaction is a major strategic priority
69% of FEMA‟s
workforce was not
included in the
First survey pool
FEMASTAT
evaluated who
was surveyed9 White men at the
GS 13-15 level
were slightly over-
represented
among survey
respondents
Respondents were
representative of a
cross-section of
Next FEMASTAT the agency‟s
analyzed the
representative programs
nature of
respondents
27. To accurately highlight negative and positive Three types of
trends, FEMASTAT analyzed scores three analysis
different ways for all 71 survey questions 9 included:
Highest & Lowest
Scores
Biggest Increases
& Decreases
Largest variance
from DHS or
Government
averages
The analysis
revealed two areas
of strength and
three opportunities
for improvement
FEMA senior
leadership reviewed
the analysis and
developed actions
to address
concerns about
leadership, fairness,
and professional
development
28. In December 2010, FEMA senior leaders decided on the
following actions:
Increase transparency by televising key senior leadership
meetings (including FEMASTAT)
Seek more employee input by expanding the Employee
Viewpoint Survey to 100% of FEMA‟s workforce
Enhance visibility of FEMA‟s professional development and
emerging leader programs
Promote fairness within the employee awards program by setting
standards for qualification, allocation and distribution
Establish a two-way “culture of communication” at FEMA
Employees are currently completing the 2012 survey
and FEMASTAT will review those results in December
2012 to evaluate the impact of these decisions.
29. In summary:
At every level of government, in every organization, you can
use data to drive decision making and to demonstrate the
impact of those decisions
The key is to (1) pick a problem you want to solve, (2) commit
your organization to solving it, and (3) demand that data be
used to inform decisions along the way
Sometimes this requires investing in the data first
“Optimism, without data, is just an emotion.”
- - Dan Tangherlini, City Administrator, Washington DC, 2007-2009
Questions?
30. REFERENCES & NOTES
Citation Slide Reference/Note
The Individuals with Disabilities Education Act (IDEA) requires states to ensure children with disabilities, to the
maximum extent possible, are educated with children who are not disabled. That ideal scenario is considered
1 2 the “least restrictive environment” under Federal law; therefore, the “most restrictive environment” is one where
students with disabilities are educated in total isolation from their non-disabled peers.
The $154 million includes tuition only, and does not include administrative costs and some expenditures on
related services. The total amount spent in the non-public tuition fund in FY 2010 was $167 million: District of
2 2 Columbia Government FY 2013 Proposed Budget and Financial Plan, Non-Public Tuition, page D-97.
http://cfo.dc.gov/cfo/frames.asp?doc=/cfo/lib/cfo/budget/fy2013/chapter/public_education_system/gn_npt_chapte
r.pdf
3 3, 4 Graphic on non-public student enrollment extracted from July 20, 2011 SPEDSTAT on Non-Public Tuition.
District of Columbia Government FY 2013 Proposed Budget and Financial Plan, Non-Public Tuition, page D-97.
4 4 http://cfo.dc.gov/cfo/frames.asp?doc=/cfo/lib/cfo/budget/fy2013/chapter/public_education_system/gn_npt_chapte
r.pdf
Department of Health and Human Services, Centers for Disease Control and Prevention, HIV Infection
5 5
Reporting web-article, accessed on July 25, 2012: http://www.cdc.gov/hiv/topics/surveillance/reporting.htm
All charts and data extracted from the District of Columbia Department of Health HIV/AIDS, Hepatitis, STD, and
TB Administration (HAHSTA) Annual Report 2011:
6 6, 7, 8
http://doh.dc.gov/doh/frames.asp?doc=/doh/lib/doh/services/administration_offices/hiv_aids/pdf/HAHSTA_ANNU
AL_REPOR_2012.pdf
7 9 Best Places to Work from: http://bestplacestowork.org/BPTW/rankings/detail/HS18
8 9 Retention data from Federal Employee Viewpoint Survey
9 10, 11 FEMASTAT Presentation on Employee Satisfaction, December 7, 2011.
31. GovDelivery
Analytic Star Potential – Go for the Extraordinary
NextGen Conference – July 2012
Dennis R. Still, Client Performance Analyst
31
32. Where would you like to fall?
AVERAGE = a typical
amount, rate, degree; norm
EXTRAORDINARY = exceptional in
character, amount, extent, degree;
noteworthy; remarkable
Trying to push ourselves to think beyond what is “average, comfortable, or easy” is
difficult. At GovDelivery, we believe that governments can achieve extraordinary levels
of engagement with their citizens. Helping them is what we do best.
32
33. How GovDelivery Works
1. Utilize existing 2. Place 3. Public selects 4. Personalized
Web and Social subscription links specific types of messages are sent
Media Content. and maximize information of automatically via
signup rates. interest they want to email, SMS, social
receive. media, & RSS
Content from Distribute content in
across Broad range of topics across multiple channels;
Website and enterprise; connect across post to social
outside Social the entire agency or level of media, enable
Media government to reach more sharing on
consumer web
Properties people
33
34. Settling for Average
City A - Average Opens & Clicks (Jan - Nov 2011)*
Average Open Rate Average Click Rate
35.25% 12.12%
GovDelivery Clients by Market (Jan - Nov 2011)*
Average Open Rate - All Average Click Rate - All
25.25% 10.25%
Average Open Rate - State & Local Average Click Rate - State & Local
24.20% 7.20%
Average Open Rate - Federal Average Click Rate - Federal
45.40% 22.40%
Average Open Rate - Europe Average Click Rate - Europe
10.00% 7.10%
*Data is not real and only presented for demonstration purposes
Findings:
City A has much better Average Open & Click Rates than all other clients
34
35. Presenting Extraordinary – When you can do this...
County # of New Subscribers # topics
Population Subscribers Per Month offered Automation Top Topics
• Sheriff
~250,000 • Bid invitations
• Employment
Launched: 6,595 150-250 82 90%+ • News
March ‟ 07 • Crime
• Public Health
• Commissioner‟ s Newsletter
~200,000 • Planning notices
Launched: 4,728 250-300 35 ~25% • Safe Kids
• Calendar of Events
April ‟ 10 • Solid Waste
• Public Library
~170,000 • Employment
• Parks
Launched: 3,211 150-200 55 ~80% • Farm Museum
July ‟ 06 • Capital Projects
• County Newsletter
• Parks
• Jobs
~90,000 • Employment
• Crime
Launched: 9,886 125-150 65 ~50% • Camping
January ‟ 07 • Taxes
• Flood alerts
• Public Health
City A • Jobs
• Tax Forfeited Properties
147,076
9,508 75-125 117 ~75% • Convictions / Ten Most Wanted
Launched: • Bid
September ‘01 • Calendar
Compare City A to other cities and counties of comparable population
size. Benchmark their performance against comparable agencies.
35
36. Presenting Extraordinary – Benchmarking cont.
Avg. New
# of Subscribers / # topics
Population Subscribers Month offered Automation Top Topics
• General Newsletter
~300,000 >500 • Arts Council
• Libraries
Launch: 158,676 (some months over 139 ~60% • Jobs
October ‟ 06 1,000) • Utilities
• Parks/Rec/Museums
• Events
~250,000 >400 • Library
• Services & Facilities
Launch: 45,007 (some months over 109 ~70% • Jobs
March ‟ 03 1,000) • Public Auction
• Elected Of f icials
~135,000 >400 • Seasonal Programs (Parks)
• Leagues / Events
Launch: 26,648 (some months over 105 ~50% • Theatre / Culture
Sept. ‟ 09 2,000) • Community Programs
• Community events
~260,000 ~20% • General newsletter
Launch: 56,884 1,000 – 2,000 203 (still adding • Elected of ficials
• Public health
February ‟ 11 automation) • Parks
• Parks
~120,000 >350 • Jobs
Launch 31,316 (some months over 150 ~80% • Emergencies
• Weather Response
August „ 07 1,000) • Neighborhood Newsletter
Now, compare City A to Top Performers within criteria. Benchmark their
36
performance against those agencies – power of influence.
37. Presenting Extraordinary:
Key Observations From Benchmark Comparison
City A performs high on many levels
• Large # of topics offered
• Topics consistent with what works in other local governments with minor
exceptions
Clear room for improvement, particularly in growing outreach base
• More recent implementations have higher new subscriber rates
• Newsletters, events, health, and parks content all do very well in top
performing local governments
Highlight what department/agency is doing well – praise those
efforts. At the same time, highlight the importance of how they
might improve based upon metrics accessed. Make it actionable!
37
38. Presenting Extraordinary – GovDelivery Network
• 114,247 new subscribers in first half of 2011
• Accounts using the network got 50% of new
subscribers from the GovDelivery Network
• The Governor of TX gets over 90% of new 17 Texas Agencies
subscribers from the GovDelivery Network use GovDelivery
1.5M total subscribers
• Los Alamos National Laboratories gets almost all
of its new subscribers via cross promotion with
NASA
• Even small percentage increases in new
subscribers via GovDelivery Network can lead to
thousands of engaged consumers of agency
information – creating mission value
38
39. Presenting Extraordinary: Strong Vital Stats 2012
Total
Impact: Subscribers
125,112
1) Agency A continues to GovDelivery
Total
Network
experience significant Subscribers
Subscriptions
growth in 605,250
55,980
subscribers, opens, and
clicks.
2) The Average Engagement
Agency
Rate for Q1 2012 for Federal Current A Total
Engagement Messages
Agencies is 52%. In the last Rate Sent
15,498,172
90 Days, Agency A’s 65%
Engagement Rate is 65%.
Total Opens Total Clicks
792,456 344,160
39
41. Presenting Extraordinary: New Tools & Impact
Tax Tips - Web App & UI - Since Launch 2/8/2012
1,000
900 61% Average of New Subscribers
coming via Web App
800
700
600
500
400
300
200
100
0
UI New Subscribers App New Subscribers
41
42. Movement from Average to Extraordinary
Examine & Understand What Data You Have
Move Toward Leveraging That Data For Better Understanding
42
43. Going Beyond Descriptive Analyses
Analysis:
Two Streams
Descriptive Predictive
Concerned with summarizing the properties of a sample of Apply mathematical theory of probability to make decisions
observations/actions – What has occurred in the past and about the likely properties of populations – What can we do
can be counted in some way to help department/agency achieve better results
Mean/Median Sums/Averages Percentages Order
Move toward this
Try to move away from just descriptive patterns – predictive model
need to get data integrity established to allow
organization to mature in terms of analyses
43
44. Conclusion
Strive for Extraordinary – Think about
impact for your intended audience, examine
what questions you want to answer (hypothesis)
Don’t Settle – For bad data or lack of
data, drive for results by examining what you
have and where you might find more
Start “Small” and Work Toward “Big” –
Big Data can be daunting at first, think about
what you can do right now with smaller
data, just start examining
44
Editor's Notes
Most popular types of datasets: geography and environment, health and nutrition, and national security and veterans affairs
1 The Individuals with Disabilities Education Act (IDEA) requires states to ensure children with disabilities, to the maximum extent possible, are educated with children who are not disabled. That ideal scenario is considered the “least restrictive environment” under Federal law; therefore, the “most restrictive environment” is one where students with disabilities are educated in total isolation from their non-disabled peers.2 The $154 million includes tuition only, and does not include administrative costs and some expenditures on related services. The total amount spent in the non-public tuition fund in FY 2010 was $167 million: District of Columbia Government FY 2013 Proposed Budget and Financial Plan, Non-Public Tuition, page D-97. http://cfo.dc.gov/cfo/frames.asp?doc=/cfo/lib/cfo/budget/fy2013/chapter/public_education_system/gn_npt_chapter.pdf
3 Graphic on non-public student enrollment extracted from July 20, 2011 SPEDSTAT on Non-Public Tuition.
3 FY 2011 spending graphic extracted from July 20, 2011 SPEDSTAT on Non-Public Tuition4 District of Columbia Government FY 2013 Proposed Budget and Financial Plan, Non-Public Tuition, page D-97. http://cfo.dc.gov/cfo/frames.asp?doc=/cfo/lib/cfo/budget/fy2013/chapter/public_education_system/gn_npt_chapter.pdf
5 Department of Health and Human Services, Centers for Disease Control and Prevention, HIV Infection Reporting web-article, accessed on July 25, 2012: http://www.cdc.gov/hiv/topics/surveillance/reporting.htm
6 All charts and data extracted from the District of Columbia Department of Health HIV/AIDS, Hepatitis, STD, and TB Administration (HAHSTA) Annual Report 2011: http://doh.dc.gov/doh/frames.asp?doc=/doh/lib/doh/services/administration_offices/hiv_aids/pdf/HAHSTA_ANNUAL_REPOR_2012.pdf
6 All charts and data extracted from the District of Columbia Department of Health HIV/AIDS, Hepatitis, STD, and TB Administration (HAHSTA) Annual Report 2011: http://doh.dc.gov/doh/frames.asp?doc=/doh/lib/doh/services/administration_offices/hiv_aids/pdf/HAHSTA_ANNUAL_REPOR_2012.pdf
6 All charts and data extracted from the District of Columbia Department of Health HIV/AIDS, Hepatitis, STD, and TB Administration (HAHSTA) Annual Report 2011: http://doh.dc.gov/doh/frames.asp?doc=/doh/lib/doh/services/administration_offices/hiv_aids/pdf/HAHSTA_ANNUAL_REPOR_2012.pdf
7. Best Places to Work from: http://bestplacestowork.org/BPTW/rankings/detail/HS188.Retention from Federal Employee Viewpoint Survey
9 FEMASTAT Presentation on Employee Satisfaction, December 7, 2011.
9 FEMASTAT Presentation on Employee Satisfaction, December 7, 2011.