Barriers & Opportunities

Date: 25 August 2009

Section 1 of the Spatial information in the New Zealand economy report (PDF, 1.64MB) introduced spatial information and highlighted a number of economic issues (including 'market failure' and the role for government), whilst Section 2 identified applications of modern spatial information technology with a range of case studies and vignettes from the different sectors of the NZ economy. From this, it is clear that the use of modern spatial information technology:

• has resulted in substantial benefits to New Zealand society
• is being used more often and in new areas
• is contributing to economic growth and competitiveness.

The work also showed a number of barriers that have held back, and are continuing to hold back, the use of spatial information in New Zealand.

This section identifies some of the main barriers and outlines significant opportunities for reducing them by making a spatial data infrastructure. The reduction strategy essentially 'glues' together existing bits and pieces of critical spatial data and the infrastructure for using it, as well as fostering the use of spatial data 'culturally' by clarifying roles, responsibilities, processes, entitlements and relationships.

Early and cost-effective interventions by government are likely to be:

• solving the technical issues about exposing spatial data to the web in a few agencies and then replicating the solutions more widely including through the use of open source software
• requiring government departments and local authorities to make their spatial data available at marginal cost (which would be zero in case of web based delivery)[1]
• providing technical elements of an SDI that other parties will not make available which could include spatial standards, metadata, and catalogue and registry functions

Section 4 of the report estimates the cost of the barriers at around $0.5 billion in 2008, including a loss of government revenue of at least $100 million. The early, cost effective interventions can be expected to go some way towards realising these potential benefits.

Nature of the barriers

This section draws on concepts of market failure and government failure to identify barriers to spatial information adding more to productivity in the New Zealand economy.  It also draws on the limited existing literature and on the anecdotal evidence gathered in preparing this report.

Market failure

Key economic concepts which are often used when making the case for government intervention include the notion of a 'public good' in the economic sense (i.e., having the property of being nonrival and nonexcludable), various types of unpriced spillovers which may be associated with public goods (i.e., positive externalities), information failures, transaction costs and natural monopoly arguments.

These concepts were explained in Section 1.3.3 of this report, where it was also pointed out that the body of literature on market failures and informational economics is extremely wide and varied.

A discussion paper by the Australian Spatial Information Industry Joint Steering Committee (2002) applied these generic economic principles to spatial information and identified the key roles of government in the spatial information sector as follows:

• Developing policy and frameworks

• Promulgating and mandating use of standards and assessing compliance

• Specifying and developing products and services to meet public interest needs

• Providing infrastructure, fundamental data and basic services

• Competitive purchasing and quality control of particular services, such as data collection, infrastructure provision and access

• Ensuring equity of discovery and access to information

• Delivering broad societal outcomes, such as inter-generational equity, effective use of public resources and protecting rights of consumers.

The report identified the key tests for involvement of the public sector as:

• a public interest need is to be met, and a public agency is sole provider, or the private sector is unable to provide a particular product or service, or based on national competition policy measures, use of a public sector provider is the most cost efficient use of resources.

Key observations

• spatial data are not a public good in the economic sense of the term until they are placed in the public domain (the data are in fact to some degree 'excludable' and are not a public good by virtue of their mere existence)
• market failure does not automatically imply that government should interfere; costs and benefits from intervening must still be assessed
  • the government also has to carefully balance private sector interests with the wider public interest, and ideally act to help deliver the wider results sought while not crowding out private initiative

• similarly, a natural monopoly situation does not mean that government must perform this function; private companies can sometimes provide some of the key spatial data services, and invest in the infrastructure necessary to provide value added servies, that might traditionally have been seen as the remit of government
  • real time GPS (or GPS augmentation) for example is delivered private sector providers in NZ (e.g., Fugro); as a consequence, the role of government in entering these markets must be reviewed very carefully

• finally, dynamics can be important and if government enters or supports a sector that is moving rapidly it should also consider its 'exit' strategy - along with the preceding comment this means that the appropriate role of government is fluid and shifts over time. Assessing the government's role requires close attention to the way the technologies are developing and how market participants are behaving.

Government & regulatory failure

The concept of government failure was introduced in Section 1.3.4 of this report, which noted that limited information, limited control over private market responses, limited control over bureaucracy, and limitations imposed by political processes are all seen as key causes of government failure in the economic sense of the term. Regulatory failure is a type of government failure, albeit with its own complexities which cannot be discussed in detail in this report.

In the context of spatial information in New Zealand, it was commented several times during the preparation of this report that the government holds large amounts of spatial data but in many cases these data are either not being shared effectively across departments (i.e., held in 'silos') and sometimes not released at all, and that there is a lack of knowledge as to what data are available where, and how one can access them. This situation indicates aspects of government and regulatory failure in the economist's sense of these terms are occurring. This has led to suboptimal data sharing within government as well as lower data use and re-use by non-government entities.

Stiglitz (2000a, p. 205) provides a number of explanations for this type of public sector inefficiency, including an absence of competition (a corollary to being the natural monopoly), the absence of incentive pay and various principal-agent problems such as the pursuit of bureaucratic objectives and high levels of risk aversion exhibited by public servants.  A range of explanations for the failure to share spatial data in New Zealand based around this concept of government failure can be offered.

Many agencies buy or generate spatial data to help them perform their own functions.  They have strong incentives to do so efficiently, but the incentives operate vertically within the organisation.  Agencies have few incentives to make spatial data available to potential users who are not seen as central to the organisations' core business.

Current policy settings require government departments to make data available at the cost of dissemination.  The cost of disseminating data though is a function of how an agency organises its business.  If data and systems are organised around meeting internal needs, the cost and practical difficulty of getting the data out to other users will potentially be high.  Local authorities are permitted to set their own policies on pricing access to data and many attempt to recover costs by selling it rather than distributing at marginal cost.

Pricing policies for access to data are particularly important.  The analysis on price elasticities at Section 1.3.5 indicates that prices (even relatively low prices) can be significant determinant of greater use of spatial data. Where pricing for public data is too high, the use of spatial data by others will be less than optimal resulting in a loss of economic welfare for the economy generally.

Incentives also operate on people working within agencies to restrict access to data. Individual and managers may be unwilling to release data because they feel its quality is not high enough or over concerns about legal liability for the accuracy and quality of data. Some data held by public agencies may be held in formats that suit the organisations' internal use of the data, but are not suited to wider use without significant reformatting.  Such reworking involves multiple private actors in work which adds little value to the organisation while it would deliver broader economic benefits.

A comment from the workshop illustrates the perception that formats indeed are an issue:

"Most (all?) New Zealand framework datasets largely ignore open standards, they have adopted a proprietary information model based on many different vendor solutions".

License arrangements may also restrict access by unnecessarily restricting reuse, imposing conditions about that are hard to apply when the data are used in web applications, or taking time to complete.

People in agencies who are responsible for data do not have incentives to create metadata that describe the data they hold because they already understand the data and how it is structured.  But without good metadata, it is difficult for others to discover the data and figure out how to use it most effectively.

The problems of data access have been the subject of considerable attention; there is potentially a way around these problems that may not involve significant additional costs for government organisations. This involves agreement on open standards and protocols which if adopted across government agencies could provide a framework for access to basic data. Such a framework has been developed by the Open Geospatial Consortium which is discussed in Section 3.2 below.

Identified barriers - literature & anecdotal evidence

The Geospatial Research Centre's capability mapping report (Park et al., 2008) identified the following list of priority issues for the NZ geospatial sector:

1. Lack of geospatial understanding
2. Problems accessing NZ data
3. Staffing issues
4. Difficulties starting / developing a business in NZ and then moving offshore
5. Access to funding
6. IT issues
7. Limited awareness / use of (internationally aligned) standards, metadata and protocols
8. Government issues

A number of these are 'generic' issues, i.e., not specific to the spatial sector alone in New Zealand (notably Points 3 to 5 on the list above). Extending the capability mapping carried out by the GRC during a recent consultancy, The Hon. Gary Nairn (2008, p. 4) identified key constraints as being the lack of understanding of the importance and benefits of geospatial information (Point 1 on the GRC list above), inconsistency in data standards (Point 7), and problems accessing data (Point 2).

As noted, one of the clearest barriers identified in the interviews and workshop for this report was the limited access to spatial data held by central and local government and other public agencies. A website (http//officialinfo.wordpress.com) maintained by two individuals in the geospatial industry records a number of attempts to obtain spatial data from central and local government agencies.  It reveals that pricing, formats and licensing are key issues for more seamless access to public data.

Work for this report also shows how barriers can be perceived by users in the retail sector (see Box 7).

---

Box 7: Barriers in the retail sector

There is limited knowledge across the retail sector of the benefits of spatial information and that there is some hesitancy amongst executives and senior managers which hampers further investment. This is compounded in many cases by the costs of investing in spatial applications both in terms of cumulative licences for data and also consultancy fees. Overall, this tends to make access to detailed analysis the preserve of larger organisations who make strategic decisions across regions and who will also benefit significantly in terms of resource efficiencies.

Smaller organisations may undertake some in-house analysis however this could take up a significant proportion of their resources, and many do not have the skills required to format and analyse the data for specific purposes. For those that do use spatial information either directly or indirectly, data quality and lack of standardisation is an issue.

Overall this suggests opportunities for government enabling access to publicly produced data via the internet, rather than current limited and costly distribution through intermediaries. Further, continued development of software and the potential increase in desktop analysis packages would enable smaller organisations to adopt the technology leading to significant increases in the use of spatial information.

Source:  Ecological Associates, based on feedback from interviews.

---

There would appear to be important economic benefits to the retail sector. Adoption of open standards through a coordinated approach by the national and local governments would appear to be one option worth exploring to reduce these barriers and realise longer term and wider economic benefits for industries such as the retail sector.

Spatial data without barriers:  an efficient & effective SDI

Another way of thinking about barriers is to compare the status quo with the ideal state; a highly functional spatial data infrastructure or SDI. While there is no single agreed-upon definition of SDI it is essentially a framework of spatial data, metadata, users and tools that are interactively connected in order to use spatial data in an efficient and flexible way. A successful SDI, almost by definition, addresses the barriers discussed earlier:

The fundamental argument for investing in spatial data infrastructure (SDI) rests in its ability to address market failures, coupled with significant returns to scale and scope. Moving towards an effective SDI is like moving from 'desktop' GIS to 'enterprise-wide' GIS, that is, there are significant benefits to be had from effective sharing and re-use of data which is why A Spatial Data Infrastructure (SDI) facilitates and coordinates the exchange and sharing of spatial data between stakeholders in the spatial data community. (Rajabifard et al., 2006)

This paper does not attempt to prescribe the platonic form of an SDI; indeed SDIs around the world might be considered a work in progress. Modern SDIs are typically built from a combination of open standards, open access to distributed data maintained at source, top down action by government and bottom-up development by users. 

The reference architecture for SDIs promoted by the Open Geospatial Consortium (OGC) is a sufficient description for the purposes of this paper.  The OGC is a consortium of over 386 companies, public organisations and universities participating in a consensus process to develop publicly available interface standards:

OpenGIS Standards support interoperable solutions that "geo-enable" the Web, wireless and location-based services, and mainstream IT. The standards empower technology developers to make complex spatial information and services accessible and useful with all kinds of applications (http://www.opengeospatial.org/ogc, accessed 31 July 2009)

The standards and principles established by the OGC are worthy of consideration by government in addressing the barriers discussed in this report. A summary is provided in Box 8.

---

Box 8: Spatial Data Infrastructure: the OGC; reference architecture

The main OGC interface specifications are:

• WMS: Web Map Service

• WFS: Web Feature Service

• WCS: Web Coverage Service

• WMC: Web Map Context

• W3DS: Web 3D Service

• CS-W: Catalogue Service for Web

The OGC has defined GML (Geography Markup Language) as the format for interchanging geographic datasets. Also ISO standards play a role. Important ISO standards is the geospatial domain are:

• ISO19110: Standard for registration of feature type definitions

• ISO19115: Metadata standard for geographic datasets

• ISO19119: Metadata standard for services van de services

• ISO19139: XML encoding of ISO19115 metadata standard

Publish-find-bind paradigm

The OGC reference architecture has the following roles:

• Provider

• Broker

• User

The publish-find-bind paradigm describes the interaction between the different roles. A data provider publishes its geographic dataset as a service (e.g. WMS or WFS). The service and the geographic dataset are being described by metadata (using ISO19115 and ISO19119 standards). The metadata is being harvested into a catalogue (broker). A user queries the catalogue for a specific geographic dataset through the catalogue service (CS-W) by specifying keywords and search area. Then the user retrieves the geographic dataset through a service and is able to use the data in its own business process.

Technology
The technical elements of an SDI can be built from proprietary components or components developed by the open source community.

Web Map Service (WMS)
The WMS is a service that returns a georeferenced map (layer). In this context a map is considered as a two-dimensional visualisation (according to a predefined style) of features and can be in common formats like jpeg.  The service does not return the actual features. By default a WMS serves one or more styles per layer. By defining a style on the client side and sending this style as part of the WMS request to the server, a map with a user-defined style is obtained. In this way thematic maps can be made. The styles used for rendering the map have to be specified in a Style Layer Descriptor (SLD) document. SLD is an XML encoding for the definition of the styles of geographic features.

Web Map Context (WMC)
It is possible to combine layers from different WMS's on the client into one single map view. Requirement is that the layers are requested with the same bounding box and output size and are within the same coordinate system. The OGC developed the WMC specification to store the definition of these map views in a XML document. Using these WMC documents is it possible to retrieve map views on another client and at a later time.

Web 3D Service (W3DS)
The Web 3D Service is a service for displaying three-dimensional data. The service returns 3D geographic elements from a specified area in VRML format which are rendered on the client. Because rendering takes place on the client, real time navigation through the 3D image is possible.

Web Feature Service (WFS)
The WFS is the service for selecting, inserting, updating and deleting features and enables geographic and attribute filtering of the features. The OGC filter encoding specification has te be used for the filter definition. It describes the XML to be used to define spatial and attribute filters. The selected geo-features are returned in GML format.

Catalogue Service for Web (CS-W)
A Catalogue service offers functionality to store (harvest) and query metadata of services and geographic datasets in a catalogue.

Source Adapted from http://www.atlis.nl/eng/ogc.html

---

Making an SDI can be considered as a national level analogue of adopting an enterprise-wide spatial system in an organisation, and productivity gains of the same order can be expected.

Figure 29: The rate of return on investment (SDI or 'enterprise-wide' versus 'silos')

Source: ACIL Tasman

Due to the size of investment initially required, the 'public good' aspects of SDIs, and the nature of some of the benefits (e.g., biosecurity, etc.) some elements of this type of infrastructure will have to be funded by government. Individual companies, especially the smaller players, won't create all the networks and capabilities for an SDI on which their competitors may 'free ride'.

As high-income economies become more and more knowledge driven, the infrastructure underpinning knowledge creation, maintenance and use also become increasingly fundamental to the functioning of the economy. SDIs are an important sub-set of these knowledge infrastructures. These emerging infrastructures go beyond infrastructure in the traditional sense of roads and buildings to include investments that increase the capability and options to create, share and use knowledge:

Whatever definition is accepted, it is clear that SDIs will include elements that are related not only purely to technology (cables, computers, software, data storage, etc.) but also to the relationships, norms and rules of exchange accepted by those involved in knowledge creation and its maintenance and use. This architecture of technology and protocols will be necessary to realise the ultimate outcome of a vibrant and innovation driven market for spatial information. A country may have supercomputers and high bandwidth internet, but if there is no process or capacity for sharing data, knowledge growth will be stymied as data still sit within 'silos'. If countries do not address this problem the market will be constrained, potential new industries will be lost to other countries and international competitiveness will suffer.

Government role in making an efficient & effective SDI

To create an SDI, a set of initiatives should ideally be launched as a package, possibly a sequential package within an overall policy objective articulated in terms of developing and SDI for the nation. This essentially requires a 'step-change' in approach, backed up with ongoing effort by government and other players to implement the changes necessary over a longer time frame. Broadly, policy implementation might be conceived of as:

• specific discrete initiatives and interventions by Government, supported by
• a policy framework that will foster a culture of support for an effective SDI in governments and industry.

Specific initiatives or interventions are proposed below to address the market failure and government and regulatory failure problems discussed earlier.  These include routinely making publicity funded spatial data available through the internet so that they can be readily re-used (this addresses the government failure issues) and implementing elements of an SDI that address the spillover and public good nature of the data that would not otherwise occur if left to the market (so addressing the market failure issue).

This section also makes some suggestions/proposals about the instruments to use. Some are designed to achieve a demonstration effect in terms of Rogers' diffusion theory.  The idea being that if solutions (especially to the government failure problems) can be devised and made highly visible, they can be replicated across many agencies.

Such demonstrations should also provide information about the costs some of the proposed interventions which could plausibly come later.  This cost information (which this study does not address) would be needed to support harder interventions like regulating to make laggard agencies participate in an SDI.

Routinely exposing its own spatial data for reuse

The analysis of barriers earlier in this section indicates that considerable value can be realised simply by exposing spatial data to the wider community of users via the internet.  This value is additional to that created by systems such a those discussed in section 2 of this report, which largely serve defined sets of data to meet the needs of a specific set of users.

By exposing data directly via the internet, it is available to anyone to use, recombine and add value in ways that cannot be predicted by policy makers or even industry analysts at the present time.  Critically, in a cost constrained environment, exposing existing data could leverage benefits ahead of more costly interventions like purchasing new data, or making other parts of the technical infrastructure for an SDI.

Two examples illustrate how government data from multiple sources can be recombined:

• GeoCommons (see www.geocommons.com) delivers basic analytical and mapping services through a web browser.
• Koordinates.com (see http://koordinates.com) provides downloads of map layers in a range of formats.

Both these services contain many layers of spatial data from publicly funded sources in the case of Koordinates largely from New Zealand.

More commonly applications can consume publicly funded data (usually behind the scenes) to provide much more specific functions like locating an address on a map or layering different sets of information over a particular property.

For data to be accessible and reusable a number of the barriers identified earlier in this section need to be addressed, specially pricing, licensing, metadata, as well as providing data in forms which maximise its use and re-use.

Pricing

The work on price elasticities earlier in this paper indicates that prices (even relatively low prices) can be a significant barrier to information re-use. The main policy settings on pricing access government held information date mainly from 1997, before access to data via the internet was cheap and ubiquitous as it is in 2009 and before the nature and value of information as a good was fully appreciated.

The settings for government departments refer to 'cost of dissemination' but as noted earlier, this cost is likely to reflect the way agencies organise their business rather than true marginal costs if dissemination was a priority.

If the institutional barriers are addressed, it is likely that the marginal cost of routinely exposing spatial data via the web could approach zero. 

There would still be some costs (say bandwidth) and some agencies would lose their current revenue from sales of data.  The loss of revenue to particular agencies would need to be addressed, but there is plenty of scope to solve that problem because gains (including tax revenues to the government) from making data free would be expected to significantly outweigh the income lost.

Licensing

Work by several government agencies including New Zealand State Services Commission (2008) and the Victorian Parliamentary economic Development and Infrastructure Committee (2009) suggests that much simpler and more permissive licensing arrangements can be used for government held information without creating additional risks to government.

The work referred to suggests that perhaps 85% of public sector information could be available under the most permissive form of creative commons license which essentially requires only attribution of the source data to its originator.  Ideally users would agree to such a license as part of the process for accessing data via the web.

Metadata

To be accessible and usable, spatial data must be accompanied by metadata that describe how the data are structured, what they describe and how they can be accessed.

The metadata need to be usable by computers and human users.  Ideally there would be a mix of highly structured and standards compliant metadata as well as tagging that facilitates discovery through search engines and the emerging semantic web.

Providing data in forms that maximise reuse

To maximise the benefits of publicly funded data agencies would make the fullest possible sets of spatial data available in relatively raw form.  If agencies attempt to operate too far down the value chain or exclude access just because data is no longer of use to the agency, they will crowd out private initiative and innovation.

This involves something of a paradigm shift from thinking spatial data and systems as products aimed at specific uses and users and towards thinking of it as raw material for innovation.

Data though must be exposed in a form that can be readily consumed by those who will add value to it, especially by the people who design and provide web applications with a geospatial component.  The forms which are ideal for consumption over the web are not necessarily the forms agencies adopt for their internal use or for the products, services they offer to core users. 

Technical format issues are beyond the scope of this report, but the almost universal feedback received in preparing it was that technical issues are relatively trivial and barriers around formats are essentially institutional in origin.

The New Zealand Transport Authority's cooperative work with web developers on its InfoConnect website described is an excellent example of effective work on data format issues. The work may be having demonstration effects too and a recent report said about the website that it:

...provides third parties and the developer community with the tools to create applications using NZTA's data.  The organisation realised it could leverage experts in the field to make its data accessible, instead of trying to do it all in-house ....... [ the website's features can ] scale from one developer, working out of his or her spare room, to big corporations wanting to use information from NZTA. 140 developers have signed up ......  applications rolled out to date include programs from the AA, MetService and a few phone applications, which are available on the iTunes App store ... (NZ Computer News 2 December 2008).

This example from NZTA also illustrates a trend taken up again in Section 3.4 (where the costs of barriers are estimated). The trend is that even without any concerted action across the whole of government agencies will quite likely get better at exposing their spatial data over time.  Another example is provided by the Ministry for the Environment making its Land Cover Database (LCDB) freely available under a creative commons licence on the Koordinates website.

This trend does not provide an argument for a 'do nothing' approach.  The estimates made in section 3.4 suggest that the greatest productivity gains can be made by concerted action across government as soon as possible.

Making the technical elements of an SDI that other players will not make

Again it is beyond the scope of this study to make specific recommendations about technical issues. Analysis of market failure does suggests some components of an SDI where government could initially focus attention.

Standards

Reasonably comprehensive and mature standards relevant to describing, structuring, and transacting spatial data already exist, largely provided through a mixture of cooperation between market participants, voluntary and government action. The two main collections of spatial standards are promulgated by the international standards organisation (ISO) and the Open Geospatial Consortium (OGC). ANZLIC also works on standards at the Australia/ New Zealand level.

Mehrtens (2009) has identified a role for government in standards on public policy grounds. Activity in this area could focus on metadata standards (because metadata is clearly a barrier), and could also focus on data models for some of the subject areas which are most critical to emergency services or to the economy like addresses and roads.

Standard data models are needed so that users of have a shared an accurate understanding of how datasets are organised and how the data they relate to objects of qualities in the real world.  There will be a role for government in coordinating and providing support for standards so that development in international standards are understood and work by government, industry, the science community and academia is ordered effectively.

Other technical elements of an SDI

The OGC reference model for an SDI described briefly in  Box 8 in section  3.2 contains components which may need to be provided by governments and are so provided in SDI implementations in other jurisdictions. They include some of the catalogue components, metadata registries, and in some implementation cases, a portal component that provides more comprehensive access to data than markets would provide.

Comprehensive access via a portal is though a not a starting point or even an essential early building blocks of an SDI.

Choice of instrument

One kind of intervention is clearly indicated: making practical implementations in agencies that break down the technical barriers to exposing spatial data to the web. The solutions should be documented, costed and support for applying them elsewhere should be provided. Potentially candidates for this sort of work include applying metadata standards, data models, and designing the stacks of software which sit between spatial databases and the web are

Free and open source software offers particular advantages in this context as it is by definition replicable without payment and the improvements and adaptations made to it are in the public domain, and as indicated in Box 8, open source options for all the technical components of an SDI are available.

Free and open source software and proprietary software however have very different characteristics in terms of up front and ongoing costs, risk, and how support can be obtained. Agencies may also see a need to use proprietary software for mission critical applications or for 'heavy lifting" tasks where highly capable proprietary software excels.

Mixtures of proprietary and open source solutions are though increasingly used in implementations of SDIs and the government should thoroughly explore how open source solutions can be employed because of the potential advantages in cost and replicability.

Solutions to the more abstract institutional barriers can be the subject of demonstration effects too. Issues like privacy, copyright, licensing, and possible liability issues around data quality are likely to be common to many kinds of spatial data.

If they can be solved once, the solutions are probably highly portable. The New Zealand State Services Commission's work on Creative Commons licensing (State Services Commission 2008) is an example of this process in action.

A harder set of interventions is also indicated and could plausibly be sequenced and designed in detail after some of the demonstration effects are evident. The sequencing issue is a critical one because the barriers being addressed are institutional. The literature on government and regulatory failure suggests that if mandatory requirements aimed at making data available were imposed before  the means to comply with them are demonstrated individuals and agencies would find ways to work around the requirements and make them ineffectual.

These harder interventions which could follow after demonstration effects have been established could plausibly include some include mixture of

• incentives on chief executives and senior management to make spatial data from their agencies available
• directing government departments to make specified spatial data available in specified formats with the most liberal possible licensing, and at zero cost
• mandating spatial standards for public agencies
• legislating for Local Authorities to make specified spatial data available in the same way
• devising policies so that research with spatial outputs funded through the public good science system is publicly available and where appropriate free.

These interventions need to be supported by much better information about the spatial sector, as noted elsewhere in this report there is very little reliable information about what is going on in the sector. Like any significant policy initiative the effect of the interventions needs to be monitored too.

Dynamic policy settings

 The spatial sector is particularly dynamic. The technology is changing rapidly as are the activities of actors in markets. This means that the logical role for government is and will continue to change rapidly too.

The changes are particular important in relation to the big ticket items like acquiring and processing large sets of data. Government agencies appear to be highly responsive at the margins if there is a clear policy imperative for instance in getting the LUCAS data needed for international obligations on climate change.

There does not appear to be any established process in place to monitor what markets are doing and set cross government priorities to fill any gaps where government acquisition of data is indicated or exit from acquiring data where it is no longer required.

One suggestion for data acquisition that emerged several times during this study was the digital elevation data needed to construct a digital elevation model.  The discussion of GPS augmentation in section 2 suggests another possible role for government in supporting a consistent approach and the need for it to be very finely calibrated.

Maintaining & developing the culture to support an SDI

An effective SDI won't be created simply by making a series of specific interventions. Many of the human and cultural components of an SDI lie outside the government or the direct effect of any interventions it makes.

There needs to be an overarching and shared policy goal under which  the other participants in the spatial sector, industry, academia, and spatial professional can organise themselves and how they interact with government.

 McLaren (2006) suggests that:

Although there is no single model for a successful NSDI [ National Spatial Data Infrastructure], all are different and reflect the varying cultural, institutional and political contexts, the following critical success factors provide guidance on how to avoid potential cul-de-sacs:

• Introduce a shared governance model amongst the stakeholders that will provide strong leadership, involvement and support. Without this, there will be a high risk that the objectives of the NSDI strategy will remain unfulfilled.
• Encourage effective partnerships among the public, private, voluntary and academic sectors. Be inclusive, wherever possible, and encourage shared services.
• Cooperate with the stakeholders governing and implementing the regional SDI strategies within the country and gain from their lessons learned.
• Sustain a strong communication strategy to ensure all stakeholder communities understand what is happening, why it is happening and how they can support the initiative.
• Create a robust business case that identifies a wide range of economic, social and environmental benefits, rather than just supporting a few stakeholder communities, and an associated performance measurement framework to monitor the benefits realisation.
• Design a NSDI that directly supports political initiatives and obtain joint funding wherever possible.
• Introduce a capacity building programme to ensure that sufficient and appropriate human resource capacity is available to create and exploit the NSDI.
• Clarify the proposed scope of the NSDI over time and obtain funding and resources to support the implementation plan.
• Provide clarity and transparency of the underlying business models to the stakeholders and simplify the licensing of geo-spatial information wherever possible.
• Intervene to change the underlying principles of the NSDI. Evolution does not always generate the desired outcomes in the NSDI arena. (McLaren, 2006)

Inherent in this prescription is the need for some certainty about what the government will do and when, and also what it won't do. This need for a clear path is essentially what was suggested by The Hon. Gary Nairn that:

...strong leadership to be taken at the Ministerial level as is occurring right around the world. It is recommended that the Minister establish an Action Agenda for Spatial Information. (2008, p. 3)

If such a clear path was established the government could facilitate, foster and encourage the co-creation of an SDI along with the other parties.

Particularly relevant here are the government agencies that deal with training and education, small business support, economic development and research funding and even immigration.

These agencies cannot be expected to develop their own detailed analyses of spatial information issues; they are however very important because they control some of the levers that affect some of the barriers identified in this report such as staff and skills shortage.

The 'cost' of the barriers & the gain from removing them

New Zealand has experienced considerable benefits from spatial information but the barriers discussed in this section represent 'missed opportunities' too.  This section sets out to estimate the costs of those barriers and the productivity gains available from removing them.

In terms of the economic modelling carried out for this report, the 'lost opportunity' consists of the gap between the two curves shown in Figure 30 - where the economy could have been in the absence of the barriers.

Figure 30: The 'cost' of the barriers - Illustration 1

Source:  ACIL Tasman

Another way of stating this that, in the absence of the barriers in New Zealand, adoption and market penetration of spatial information could have been more widespread.

This is reflected in Figure 31, which places the "real", i.e., observed, adoption curve (with barriers) next to a notional adoption curve depicting the situation without barriers. At any point in time, i.e., along the x-axis, uptake would have been higher in the absence of barriers. Alternatively it is possible to examine a point along the y-axis to derive how much earlier a given level of adoption might have been achieved.

To enable the economic modelling discussed in Section 4 it was essential to gauge the likely extent of these differences. Sector-by-sector discussion was included in Section 0 of this report. The general equilibrium modelling described in Section 4 of this report was then employed to estimate the implied dollar cost of having had the barriers in place.

Figure 31: The 'cost' of the barriers - Illustration 2

Source:  ACIL Tasman

The impacts discussed in Section 0 were greatest where the barriers are a major impediment to uptake and least (even zero) where a sector already has a high level of uptake.

The "tipping" point for adoption

In general, one can say that 10 to 25 per cent of potential adopters need to have taken up a technology before the tipping point and thus the 'steep' part of the adoption curve is reached - when the 'coefficient of imitation' really kicks in (see discussion in Section 1.3.6) - that is, before critical mass in terms of demonstrating the benefits to the average user has been reached. The discussion in Section 2 of this report indicated that probably somewhere between 500 and 1,000 organisations have taken up GIS in New Zealand.

The latest available NZ Statistics figures indicate that there are over 22,000 enterprises with more than 10 employees in New Zealand - indicating that the current level of take-up is still less than 5 per cent of enterprises (and if smaller organisations were taken into account this would put NZ firmly in Rogers' 'innovator' stage and therefore at the relatively flat, initial section of the S-curve).

This means that it is more likely that New Zealand is still some distance from the steep "tipping" point on the curve. The fact that New Zealand is some way off the tipping point does in fact suggest that there may still be significant scope for government to contribute to accelerating uptake - that is to say, to bring forward the time until the tipping point is reached. Investment in a modern SDI is exactly the means by which this 'brining forward' could be achieved.

An important corollary to this is that any benefits thus generated should only be counted for as many years as the 'time to take-off' can be estimated to have been brought forward. So even if a $100 million investment in SDI removes all barriers, additional benefits attributable to the intervention cannot be counted indefinitely into the future - as the lower curve on Figure 31 shows, New Zealand would have got there eventually.

How far does New Zealand lag behind comparable countries?

In order to better understand where New Zealand might have been in the absence of barriers to adoption, a comparison of the NZ experience with spatial information related developments overseas is informative:

• In the US, a 1994 presidential 'Executive Order' launched the US spatial data infrastructure (SDI) initiatives; in Europe, ideas for a European SDI were put forward from 1995 onwards under the 'GI 2000' umbrella. Formalised national approaches from overseas now include the National Spatial Data Infrastructure (NSDI) initiative in the United States, the Canadian Geospatial Data Infrastructure (CGDI) and the Australian Spatial Data Infrastructure (ASDI)
  • While New Zealand has the broad Geospatial Strategy which was formulated in 2007, it does not as yet have a formal policy on SDI (Cottrell, 2009). In New Zealand a "federated" model has been preferred to SDI (Cottrell, 2009);

• In Australia, for example, the Spatial Industry Business Association (SIBA) formulated the Spatial Information Industry Action Agenda Positioning for Growth in 2001 to support the emerging spatial information industry. NZSIBA was only launched in 2009.
• Understanding the sector - early spatial industry 'Census' type initiatives happened in Queensland in 2002 and Victoria in 2005 (fivenines consulting, 2005) and earlier abroad. The closest NZ equivalent is the GRC capability mapping report for 2007 (Park et al., 2008)

New Zealand's relative position in terms of the way spatial information is used is much more difficult to assess. Given that there are many areas of application, there will be some in which New Zealand is way ahead (e.g., Landonline) and some in which it is significantly behind, notably because of the market's relatively small size and the limited pool of skilled staff. It is therefore hard to know whether New Zealand 'punches above its weight' or not. However there are some notable omissions where one would have perhaps expected more action by now:

• GIS/GPS in schools: some notable, privately led efforts are underway in New Zealand (e.g., Intergraph's programs); however, for comparison, GIS in schools has been happening since 1996 in Australia and ESRI's schools program has been implemented in over 500 schools. More importantly, in addition to private sector effort, there have been concerted government efforts such as Queensland Education's grants programme and other initiatives (McInerney, 2005, 2006). We are not aware of similar government programs in New Zealand
• GIS in public health and epidemiology: sentinel papers here can be traced to the mid-1990s (e.g., Hjalmars et al., 1996, Nicholson and Mather, 1996) but in New Zealand, peer-reviewed published literature has only begun to emerge in the last five years (e.g., Hales et al., 2003, Brabyn and Barnett, 2004, Hoque et al., 2004, Scoggins, 2004, Brabyn, 2005, Brabyn and Beere, 2006)
  •  
    • This does not necessarily imply that NZ did not benefit from use of the technology as most NZ grants/projects via FRST, for example, were aimed at reports for stakeholders or were internal to the department concerned[2]

• It is also possible to engage in some preliminary analysis of industry development by comparing the (admittedly limited) data from New Zealand to those collected abroad. A ready example is that of Victoria where the aforementioned census of the core spatial industry - excluding players in the government sector - indicated that around 223 spatial industry businesses generated revenues estimated at $470 million and employed some 5,570 staff.
  • Victoria's population in 2005, when the survey was undertaken, was 5.0 million, whilst New Zealand's was 4.2 million in 2007 when the NZ capability mapping occurred which estimated a turnover of around $100 million.
    • Even ignoring the fact that the NZ estimate includes government players, this shows means turnover per head of population was around four times higher in Victoria than in New Zealand ($94 versus $24)

Feedback received from a number of experts during the preparation of this report indicates that in a range of areas New Zealand is perceived to have fallen behind after a strong start in the late 1990s. Responses included comments in specific areas such as "at least two years behind" and "progress on a well thought through data strategy back at least 3-4 years".

The discussion on the price elasticity of demand suggests that a reduction in price alone could have had a significant impact on uptake over time. There is general agreement in the literature that demand is likely to be elastic (elasticity > 1). An effective 100 per cent reduction in price would be expected to raise the quantity of spatial information demanded by more than 100 per cent (this refers to a move along the demand curve, not a shift in the demand curve). How this would have fed through to final use of spatial products in the general community is however open to conjecture. It might be argued that the capability to employ spatial data is still not in place for the average citizen.

As technology is improving rapidly and more and more households and smaller companies routinely employ ICT, the extent of the lost opportunity also increases - not opening up access in 2003 is an entirely different proposition to not opening up access in 2009, because the demand curve for spatial information has itself shifted outwards. This means that pent-up (unrealised) demand is highly likely to be increasing as time passes. At the same time, increasing incomes will have an exogenous (independent) demand shifting effect.

3.4.4   Dollar estimates

Using the technique described, the cost of the barriers is estimated at around $0.5 billion in 2008. The implied loss of government revenue is at least $100 million.[3] For the reasons outlined above these figures (annual costs due to barriers) are likely to rise with each year that passes until the barriers are dismantled. The estimates presented in this report can therefore give an indication of the likely future benefits of making a concerted effort along the lines discussed in Section 3 above, most notably investing in making data accessible and reducing access charges.

The New Zealand economy as a whole can likely gain $0.5 billion or more in additional annual productivity benefits with an effective SDI in place. Even if the government took a 'self interested' view and excluded all the wider benefits from its consideration, additional government revenue from higher productivity alone would highly likely exceed $100 million per year.

To put the size of the investment and possible returns into context, if a $100 million investment in SDI brings the time to 'take off' forward by one year only, the broad benefit-to-cost ratio for the New Zealand economy as a whole and in terms of productivity only would be 5:1 (for the New Zealand government 1:1); in the event that it brings 'take off' forward by ten years, it could be 50:1 for the New Zealand economy as a whole (and 10:1 for the government). It must be stressed, once again, that this refers to pure productivity benefits and does not include all the other non-productivity benefits which are also very important drivers of investment in the sector.

Exact SDI costings were not scoped in this report, but the likely impact is likely to be somewhere in between these two extremes (i.e., higher than 5:1 but lower than 50:1 for the New Zealand economy), especially if the SDI intervention can be implemented at a cost of less than $100 million.

Future directions for the sector

In this report, the shift from single use to multiple use of spatial data has been noted under a number of different guises. Shifting from desktop to enterprise-wide use of GIS, enabling more effective 'sharing', moving towards an all-of-government approach, even the move from the so-called aggregator model to the distributed model essentially entails the same thing - spatial data are re-used, value is added in new ways, and learning becomes cumulative rather than once off. These shifts are occurring at several levels as new technologies are developed and existing ones are being harnessed better.

The most important trend occurring today is the melding of GIS and the web.  We are seeing the work of GIS professionals able to be leveraged using web technologies so that the reach of their work is increased.  The ultimate expression of that is the reach into the consumer GIS world where analysis results can be published out to consumer products such as Google Earth and Virtual Earth. (Interview, Eagle Technology)

Such changes should ultimately have an impact on the uptake of spatial information by individuals, as people increasingly see spatial technologies being employed in their homes, cars, and so on. Such shifts will in turn result in a greater understanding and acceptance of spatial information technologies in the workplace, thus removing a major barrier to enterprise-wide GIS implementation.

Enterprise GIS has become more modular recently and this modularity will continue.  This will allow more incremental implementations of GIS - adding capability to an existing system will be much lower risk than has been the case in the past.

It is very likely that further ICT advances will allow GIS users to take advantage of higher resolution content, geo-referenced video and more frequent collects of data.

A key feature of Web 2.0 is the notion of web applications being able to post information back onto the server.  It is likely that this will result in more consumer-driven corrections of government data sets as citizens 'red line' errors back to the producer.

The convergence of mobile phones and GIS offers significant national benefit.  One example is the offering of traffic congestion information. (Interview, Eagle Technology)

Another interesting direction for the future is linked to e-governance and issues of devolution or decentralisation. Spatial information technology can enable communities to understand their environments better, and assist in communicating and coordinating 'grassroots' action. In planning tourism for the Golden Bay region, for example, GIS provided a tool in resolving the conflict between industry interests and the community's interest in preserving the relatively undeveloped beaches. The community collected and visualised their information and this gave the local residents a stake in decision-making and planning.

There is a vision of a future where anyone can 'mash up' maps in an easy operating environment, say, a tool that is the equivalent of word processing software that is now part of standard software uploaded on every PC. At present, it looks like that is a few years away from becoming a reality; however, in the longer term it is entirely conceivable that various technologies integrate spatial information:

• GPS/GNSS mobile telephony applications that go beyond simple mapping (e.g., tracking epidemics; direct marketing and so on)
• computer games and consoles
• more advanced navigation systems (potential accident avoidance is in-built)
• integration with credit card chips or other handheld devices
• small business management systems (integrating time, spatial, and price dimensions for stocks & inventories, scanning sources in real time to permit coordinated reordering that maximizes efficiencies, etc.)
• night vision and other types of equipment to assist the visually impaired
• car sharing schemes or other travel planning assistance

more advanced planning and development processes integrating information on existing quality and rate of decay of assets and roads as well as public housing stocks, etc., to ensure projects are timely and objectives are met efficiently, and so on.

Figure 32 Golden Bay Participatory GIS Project

With an effective distributed system underpinned by SDI there is almost unimaginable potential for spatial information to become 'mainstream' - from enterprise systems to government, consumers and in social networking applications:

• government applications in emergency management, security, biosecurity, social services, policy and planning, service delivery, transport management, etc.
• management of climate change and water resources along with natural resource management generally
• business systems including facilities management, logistics, planning, marketing, retail management, consumer services, etc.
• social networking including consumer communications, product research, consumer feedback, social inclusion.

Predicting where spatial information technologies will go is impossible, precisely because of some of the option characteristics inherent in spatial data and the new technologies that are developing and blurring the boundaries between existing technologies. Processes will be influenced by the way in which technologies are converging. Some idea of the value already being recognized is in the options that companies are taking out in the acquisitions of spatial companies recently. Examples include:

• Pitney Bowes acquisition of MapInfo
• Nokia's $9 billion acquisition of Navtech
• Gamin's acquisition of Tom Tom
• Quascos acquisition of Century 21 in Australia.

It is also understood that Microsoft is acquiring businesses with spatial capacity. These investments indicate that it is a sector with a future.

The role and future of spatial information is also intricately and importantly linked with New Zealand's wider innovation system. The speed with which modern spatial information technology applications are developed and spatial information permeates the NZ economy is at least partially a reflection and measure of the capability of the innovation system; at the same time, this capability is affected by modern spatial information technology.

As noted during interviews carried out for this study, New Zealand is certainly perceived as a place where innovative activity is encouraged, and the 'core' spatial information innovators in the private sector are particularly creative, knowledge intensive technology transformers. Spatial information, in short, has the potential to play an important role in the NZ economy in future.[4]

 

---

References

[1] Noting that many TLAs suffer from capacity constraints that would make it difficult to comply with this requirement in the short term it is clear that any such intervention would have to be sequenced and funded appropriately.

[2] Pip Forer, pers.comm. 15 June 2009

[3] This result is not 'self-evident' from the results of the economic modelling shown in Section 4 of this report, where the economic gains are split into 'productivity' benefits as well as government tax and tariff revenue (the latter increase by $54 million between the two scenarios shown); however, given that real GDP is estimated to grow by around $0.5 billion with the removal of barriers, based on the historic shares of private and government consumption in total expenditure, the government share of the gain would likely exceed $100 million. If transfer payments administered by government were included (e.g., pensions) this figure could be even higher.

[4] The possibility of foreign takeovers/acquisitions of NZ spatial innovators, such as occurred in the case of Navman Wireless in 2003, must also be noted.