Sustainable Public Financial Management

Doug Hadden, VP Products

Make sure you calculate the Return on Investment of open data projects. Experts warn us: the open data business case is often lacking. Traditional ROI calculations are too narrow.

Traditional ROI Calculation and “Government as Platform”

A traditional calculation compares open data costs and lost revenue from selling the data relative to additional tax revenue achieved. Tim O’Reilly  coined the notion of “government as platform” where open data (such as GPS) generates significant economic development. Some, like Andrea di Maio, disagree that governments should be acting as platforms.  Yet, this is what governments do: build economic platforms such as highways and bridges.

The problem with:

$TaxRevenue > $OpenDataCost + $DataSales

is that the government organization that provides data and benefits directly from selling the data to the private sector does not directly receive the additional taxes nor can these taxes be easily directly attributed to open data.

1. Economic Development and “Government as Platform”

Open data as policy can be compared to stimulus programs. Stimulus programs can improve employment and reduce bankruptcies. This reduces social services costs:

$TaxRevenue + $SocialServicesCostAvoidance > $OpenDataCost + $DataSales

2. Freedom of Information Costs

David Eaves has pointed out that open data can reduce the costs of Freedom of Information:

So in a world without an open data portal the hypothetical cost of fulfilling these “Canadian” downloads as formal access to information requests would have been $967,184.46 in January alone. Even if I’m off by 50%, then the cost – again, just for January – would still sit at $483,592.23. Assuming this is a safe monthly average, then over the course of a year the cost savings could be around $11,606,213.52 or $5,803,106.76 – depending on how conservative you’d want to be about the assumptions.

Open data portals can eliminate the need for many Freedom of Information requests – a high cost to governments.

$TaxRevenue + $SocialServicesCostAvoidance + $FOIAvoidance 
> $OpenDataCost + $DataSales

3. Productivity

Open data shines light on the public service. Although transparency does not necessarily mean accountability in the sense of enforcement, it has the power to change behaviour. Government organizations become aware that the public is watching. Civil service behaviour changes.

$TaxRevenue + $SocialServicesCostAvoidance + $FOIAvoidance + $ImprovedProductivity 
> $OpenDataCost + $DataSales

4. Corruption

Procurement expert Jorge Claro estimates that procurement corruption costs up to 20% for governments in developing nations. That’s why countries like Timor-Leste open budget and procurement data: to improve oversight by vendors and civil society to reduce corruption. (This also reduces the cost for auditors when the press is examining all large government purchases.)

$TaxRevenue + $SocialServicesCostAvoidance + $FOIAvoidance + $ImprovedProductivity + $ReducedCorruption

> $OpenDataCost + $DataSales

5. Trust

There remains a distrust of government in many countries. Not just in developing countries: distrust of government is a powerful narrative in American politics. The distrust in government can boil over to protest (Tea Party, Arab Spring, Occupy Wall Street etc.) This costs governments – although the release of data might give cause for protest in some cases!

$TaxRevenue + $SocialServicesCostAvoidance + $FOIAvoidance + $ImprovedProductivity + $ReducedCorruption +$ReducedProtests

> $OpenDataCost + $DataSales + $IncreasedProtests

6. Tax Compliance

Developed and developing nations alike struggle with tax avoidance. This is especially rampant when citizens believe that governments are wasting tax money. This is often made acute because of the lack of open data. There are frequent studies that show a significant delta between how governments spend and what citizens think governments are spending. In particular: foreign aid and crime. One can make the argument that Scandinavians are less concerned about high tax rates than Americans because they see value.

$TaxRevenue + $ReducedTaxAvoidance $SocialServicesCostAvoidance + $FOIAvoidance + $ImprovedProductivity + $ReducedCorruption +$ReducedProtests

> $OpenDataCost + $DataSales + $IncreasedProtests

7. Reduced Perceived Business and Donor Risk

Transparency is fundamental in many governance valuations used by the World Bank World Governance indicators (WGI) and the Millennium Challenge Corporation (MCC). Transparency indicators used to demonstrate reduced business risk and help generate donor funds in developing countries include Open Budget Index for open budgets and Revenue Watch Index for revenue transparency from extractive industries. International transparency standards include the Extractive Industries Transparency Initiative (EITI) and the International Aid Transparency Initiative (IATI). Public Expenditure and Financial Accountability (PEFA) assessments are gaining widespread use by donors in making funding decisions. Transparency is a key element for 6 measurements in the PEFA Performance Measurement Framework:

Transparency can reduce perceived business and donor investment risk in developing countries. This increases taxes, improves outcomes and ultimately improves credit ratings.

8. It’s a Network

ROI in the physical world is a diminishing returns calculation. Each new market for toothpaste increases costs. The virtual world is one of increasing returns. Each new chunk of open data adds value to previous chunks of open data. And the costs to collect and maintain open data goes down because the infrastructure scales. (That’s why Amazon can provide such value for the Elastic Cloud.)

∑($TaxRevenue + $ReducedTaxAvoidance $SocialServicesCostAvoidance + $FOIAvoidance + $ImprovedProductivity + $ReducedCorruption +$ReducedProtests)↑increasingreturns

> ($OpenDataCost + $DataSales + $IncreasedProtests)↓reducing costs

Conclusions

The ROI calculation is a bit mangled here. Yet, there is a compelling value proposition for government organizations to take the open data journey. The calculus depends on the country situation – in some cases open data can pay for itself thanks to reduced corruption or protest.

Doug Hadden, VP Products

Government “shared services” are meant to save money and improve agility. One option is to share a data centre running Enterprise Resource Planning (ERP) software. The UK National Audit Office  survey (Efficiency and reform in government corporate functions through shared service centres) of 5 ERP shared services in the UK government shows that this approach increases costs:

• £1.4bn spent to date on five shared service centres
• £159m of planned savings by end 2010-11
• £255m is the actual net cost of those shared service centres tracking benefits

ERP or GRP?

FreeBalance is a provider of Government Resource Planning software. No private sector “enterprise” software. There is mounting evidence of the lack of ERP success even in the private sector. More so in the public sector because the software was originally designed for the private sector in mind.

Most critiques of ERP failure in government point to project management causes. My sense is that Project Management 101 will not solve these problems, although there is a governance problem.

There is something fundamentally wrong about with ERP in government

The 5 ERP systems were from the big two providers. The primary problems with ERP are:

• Inappropriateness of these solutions in government increases complexity because of needed customization: “The software systems used in the Centres have added complexity and cost. All the Centres we visited use Enterprise Resource Planning (ERP) software systems. These are complex and have proven to be expensive.”
• High cost of implementation and  maintenance: “, the cost to establish, maintain and upgrade these systems is high.”
• Lack of agility to handle changing legal reform and government mandates: “these ERP systems are complex and it is not easy to modify them when needs change, such as when an organisation is restructured or processes are redesigned.”
• Unnecessary system footprint: “we found the Centres are only using a small part of the capability their ERP systems provide.”

In my opinion, this notion that ERP fails to achieve expectations in government because systems are ”overly tailored” is a crock. Government organizations have unique mandates. There are far more ‘lines of business’ in government than a private sector conglomerate. The inability to easily adapt to meet government needs is a weakness of private sector ERP applications.

The Governance factor

Governance is a critical factor. The report states: ”most shared service customers do not have adequate information on costs, performance and benefits to make informed decisions.” The NAO analyze some of the factors such as: “smaller departments and agencies have been reluctant to share services with a large department as they feel they will not get the same attention as the primary parent customer.”

My view is that improving the governance structure and providing the voice of the users is important, but not critical to success. What the NAO seems to miss – and pretty much anyone who analyzes ERP failure in government - is that the most important stakeholder to ensure shared services success is at arms’ length – the software vendor. That’s right – governments are expected to consolidate, customize and maintain highly complex ERP software without manufacturer commitment to support government needs. Manufacturers should be contractually obligated to meet government needs – no customization by the customer. Where code developed by the manufacturer is fully supported by the manufacturer.

Simpler Solutions?

I often get the reaction from those people who have used FreeBalance and one of the big ERPs: “FreeBalance is simpler and easier to use.”  The NAO suggest that government organizations consider “smaller, simpler software solutions.”

It doesn’t always sound like “simple” and “small” are necessarily good characteristics. It implies that there is something missing. “Optimal” might be a better term. Optimal as in:

• No private sector functionality increasing the technical footprint (amount of code, number of tables)
• Designed for government use so is intuitive to government users
• Use of modern computing architecture, so no client/server base code to reduce performance

FreeBalance staff celebrated World Read Aloud Day, March 7th. in 14 countries and 9 languages.


Created with Admarket's flickrSLiDR.