EENABLING CENSUS 2006

1
E-ENABLING CENSUS 2006

• Brian Pink
• Government Statistician
• Statistics New Zealand

2
SO WHATS THE BIG DEAL?

• After all we can already do lots of things over
  the net
• Banking
• Newspaper delivery
• Travel Bookings
• Researching and buying products and services
• Video-conferencing
• Accessing statistics
• Also growing range of e-govt applications have
  been e-merging (tax lodgement,company
  registrations, etc)

3
Putting Theory into Practice

• Background on the Census
• Census Logistics
• Statistical Challenges
• Reflections on the Theme
• Some Generic Lessons

4
Background

• Held every 5 years
• Legal obligation to participate
• Involves every household AND
• Every individual in every household
• Last held on 6 March 2001

5
Census Logistics

• Planning commences in 2002
• In principle decision to use internet late 2002
• Agreement to work with Canada, Australia on
  internet option - end 2002
• Content reviewed and finalised subject to testing
  - Oct 2003
• NZ field testing for internet option started Mar
  2003
• Further field test over 2004 in all 3 countries

6
Census Logistics

• Final dress rehearsal for next NZ Census in Mar
  2005
• final questionnaire used
• all systems, processes and field enumeration
  procedures fully tested
• Forms finalised and printed over May Sept 2005
• Census conducted in March 2006
• Around 7,000 enumerators and field managers
  recruited and trained Nov 2005 Feb 2006
• 6 million forms distributed Nov 05 Feb 06
• 180 processing centre staff recruited and trained
  Feb Apr 2006

7
Statistical Challenges

• Will the same people who normally complete census
  schedules, both individual and the household do
  so in 2006?
• If something changes (ie more kids want to do
  their own form on-line) what impact will this
  have on data quality?
• How can we manage over/under enumeration (eg
  someone lodges electronicially early but is
  included again when staying at a friends place
  around Census night)

8
Statistical Challenges

• Will the distribution of the population be
  impacted by electronic lodgement (including
  people going overseas)
• How can we avoid people incorrectly claiming to
  have lodged electronically when paper forms
  collected from household by the enumerator
• Will peoples responses be influenced by
  perceived anonimity of e-lodgement

9
So Why is an E-enabled Census such a big deal?

• Technologically, it probably wont be
• Need authentication mechanism
• Need secure transfer
• Need to support multiple sessions to complete
• Need to support high transaction loads at peak
• AFTER ALL, THERE IS NOTHING NEW IN ANY OF THIS
• Logistically could be a nightmare if we havent
  got all aspects integrated with our traditional
  method
• Need strong and fast links to and from field
  enumeration workforce

10
So why is using the internet such a big deal?

• Census model of field enumeration has remained
  largely unchanged over the past 50 years or more
• Refinements made each time but basic model
  unchanged
• Internet lodgement option overlays a whole new
  level of complexity
• Keeping track of which households and individuals
  within households are opting for which response
  option
• Advising enumerators quickly of electronic
  receipts

11
So why is using the internet such a big deal?

• Census field costs are based around defining
  limited collection areas with hand delivery and
  preliminary count of no. of forms delivered
• Call back to pick up and check no. of forms
  offered matches expectations (or explanation) -
  this is a simple but very important quality check

12
Some generic lessons

• E-enabling is often not about the technology(s)
• Its the business processes both the new and the
  old and the means of integrating them that is
  often the biggest challenge
• Be certain that you understand the implications
  of the changes you are making - the consequences
  are not always positive (UK Census 1900 release
  last year)
• In cases such as Census, there is no second chance