In an article today, the Guardian Australia outline a flawed critique of the Institute of Public Affairs’ latest research report The Growth and Complexity of Environmental Regulation.
The report is based on a quantitative analysis of the Environment Protection and Biodiversity Conservation Act 1999 conducted using RegData. RegData is a method of counting the number of regulatory restrictions in a body of legislation. It was developed by the Mercatus Center at George Mason University in the United States, and together with researchers from RMIT University and the IPA was expanded to include a count of Australian regulations at the state and federal level.
The Guardian put forward two main criticisms of RegData, which the report uses to demonstrate that since the year 2000 there has been a 445 per cent increase in environmental regulation enabled by Australia’s primary environmental legislation.
The first is that RegData does not capture the efficiency or effect of regulations. This is a misrepresentation of what RegData was designed to do. As the IPA explained to The Guardian when contacted for comment: “RegData isn’t designed to measure the effectiveness of regulations, and doesn’t claim to be – it’s designed as a measurement tool to count them. We make no claim about the effectiveness of regulation based on RegData, we simply note that there are so many of them that they become difficult and costly to comply with.”
This was also explained to The Guardian in an email from Patrick McLaughlin, the creator of RegData. He said: “RegData is a way to measure, like a scale. When you step on a scale, you get a number. That number doesn’t necessarily differentiate muscle, fat, and bone. But it’s a very useful number nonetheless.”
The second criticism is that RegData is a “political project” that is “ideologically driven”. As the IPA explained in the email response to The Guardian, this is completely untrue. RegData is simply a dataset that provides a consistent and objective count of regulations – the dataset does not make a claim about anything, it is simply a tool to allow for more accurate research into the effects of regulations.
This criticism demonstrates that The Guardian does not understand how RegData works and what it is designed to do. By combining RegData with other datasets, such as employment statistics and GDP figures, researchers can quantify the effect of regulation on the economy. Contrary to claims in the article that RegData caters to “a neoliberal, anti-regulatory ideology”, it has been used in research from left-leaning think tank the Brookings Institution to claim that increased regulations have not had an adverse effect on productivity.
A technology that can be used by all sides of politics is clearly neither a “political project” or “ideologically driven”.
The Guardian article also fails to mention that RegData has been used in countless academic articles published in respected journals, such as The Quarterly Journal of Economics (the oldest professional journal of economics in the English language, edited by Harvard University’s Department of Economics), Review of Economic Dynamics and Public Choice.
RegData has been referenced by a number of respected publications such as The Economist, The New York Times, The Wall Street Journal, Forbes, Politico, The Hill, and Financial Post. Additionally, it has been referenced by The Executive Office of the President of the United States, The State of Ohio, and The State of Illinois. A number of jurisdictions across North America have used RegData to measure their regulatory reform efforts, including Idaho, Missouri, Kentucky, and Iowa.
Despite claims to the contrary in the article, a significant and growing body of research demonstrates the costs that regulation imposes on society. The accumulation of regulation stifles economic growth, prevents investment and innovation, reduces wages, and has a disproportionate impact on small businesses, unskilled workers and low-income households.
It is also worth noting that in the 2018 critique cited by The Guardian, Professor Jodi Short acknowledges that RegData is “a particularly important development in the regulation counting project” and that “RegData is poised to gain more widespread use by governments implementing regulation counting policies and thus must be taken seriously not only as an intellectual project, but also a policy tool.”
Additionally, it is worth noting that an academic article co-authored by Patrick McLaughlin which introduces RegData was published in the journal Regulation & Governance which was edited by Professor Short at the time.
RegData is recognised as an important and effective tool that has furthered the field of regulatory economics. It is open source, meaning that anyone can access the data free of charge. This is the basis of open and free academic inquiry and leads to a higher quality public debate. The IPA will continue to implement RegData into research to capture the effects that regulation has on the Australian economy.
Full email exchange with The Guardian Australia:
From: Evan Mulholland
Sent: Wednesday, April 29, 2020 5:41 PM
To: Graham Readfearn <[email protected]>
Subject: RE: RegData and “green tape”
Thanks for your enquiry and giving the IPA an opportunity to respond.
On your first point: UC Hastings law professor’s critique of RegData is outdated. It was published in 2018, and since then there have been a number of revisions to RegData. The latest version of RegData (3.2) explicitly addresses the criticism that it does not differentiate between the use of key words. A history of RegData can be found here: https://www.quantgov.org/history
The foundation of RegData is Natural Language Processing. NLP is quickly being recognised as a leading method of analysing regulations, see for example this paper https://www.nber.org/papers/w26856.
All code used and data collected in the RegData methodology is subject to quality review, including code and data used in the IPA report.
On the argument about effectiveness, RegData isn’t designed to measure the effectiveness of regulations, and doesn’t claim to be – it’s designed as a measurement tool to count them. We make no claim about the effectiveness of regulation based on RegData, we simply note that there are so many of them that they become difficult and costly to comply with. Also, linking RegData with other datasets allows for an analysis of the effectiveness of regulations. And RegData is a completely open source platform so anyone can go and do this themselves if they’re unhappy with our analysis.
This is what we say in the IPA report:
This paper builds on that research by using a new technology, RegData, which was developed by leading researchers at the US-based Mercatus Center at George Mason University to measure the extent and growth of regulation. RegData uses machine learning and textual analysis to count the number of ‘regulatory restrictions’ contained in a given piece of legislation or subsidiary legislation.
Regulatory restrictions are clauses in legislation and subsidiary legislation which compel or prevent certain behaviours. There are five such clauses captured by the RegData methodology: “shall”, “must”, “may not”, “prohibited”, and “required”. These words constitute what this report refers to as “regulation”. Subsidiary legislation refers to the documents created by the relevant minister or department that impose additional legal obligations to properly implement the Act, as authorised under the Act.
On the note about the motive of RegData being a “political project” and “ideologically driven” we would note that RegData has been referenced in academic articles in reputable journals such as The Quarterly Journal of Economics (the oldest professional journal of economics in the English language, edited by Harvard University’s Department of Economics), Review of Economic Dynamics, Public Choice, and a range of papers published by the National Bureau of Economic Research and the Brookings Institution (Two Brookings papers: https://www.brookings.edu/wp-content/uploads/2017/09/2_gutierrezphilippon.pdf, https://www.brookings.edu/wp-content/uploads/2017/03/1_fernaldetal.pdf).
Another point on claims of RegData being a “political project” and “ideologically driven” Brookings Institution, a left-leaning think-tank in their paper uses RegData to claim that regulations have not had an adverse effect on productivity. This clearly demonstrates that RegData is not a “political project” or “ideologically driven” – it’s just a measurement tool.
We argue there are clear links between the level of regulation and economic performance. This academic article in the Review of Economic Dynamics uses RegData to demonstrate that US GDP would’ve been 25 percent larger in 2012 if regulations remained at 1980s levels: https://www.sciencedirect.com/science/article/pii/S1094202520300223?via%3Dihub.
Google Scholar link to research citing RegData: https://scholar.google.com/scholar?oi=bibs&hl=en&cites=9770361387012779911,8506606037394386743,6947330530897730466
RegData has also been referenced in The Economist, The New York Times, The Executive Office of the President of the United States, The State of Illinois, The State of Ohio, FactCheck.org, The Wall Street Journal, Forbes, Politico, The Hill, and Financial Post.
There are a number of jurisdictions in North America – which have applied the RegData method as a part of their red tape reduction initiatives. Such as Idaho, Virginia, Missouri, Kentucky, and Iowa. There are nine States with passed legislation/Executive action taken, four States planning to introduce, three States actively considering and six States With attempted legislation
The IPA report also analyses the number of pages of legislation administered by the federal environmental department, which finds that there has been an 80-fold increase in Commonwealth environmental law since the first Commonwealth-level environment department was established in 1971.
If you want any comment from the IPA on our research, the following can be attributed to Institute of Public Affairs Director of Research, Daniel Wild:
“Red tape is a critical issue, which IPA research has found costs Australia $176 billion in lost economic output each year.”
“Recent IPA analysis found that legal activism by green groups enabled by Section 487 of the EPBC Act has put over $65 billion of investment at risk in Australia by holding major projects, such as dams and coal mines, up in court for a cumulative total of 10,100 days since the year 2000.”
“Any economic reforms post coronavirus must focus on cutting taxes and cutting red tape. The IPA supports the Morrison Governments aim of a business-centric recovery enabled by ‘aggressive deregulation’ and tax cuts.”
“RegData is a revolutionary approach to measuring and cutting red tape which has had tremendous success in Canada and the United States. Governments across Australia should incorporate the RegData method into their red tape reduction and economic reform agendas.”
“RegData can be used to measure the red tape burden, track government progress in reducing red tape, compare the regulatory burden across industries and across states and identify areas of prioritisation for governments.”
“RegData measures the extent and growth of regulation over time, and thus provides a database to measure a range of matters including the cost impost or effectiveness of a given set of regulations.”
Director of Communications
Institute of Public Affairs
The Guardian | Australia
On Wed, 29 Apr 2020 at 16:36, Evan Mulholland <[email protected]> wrote:
Working on a response which we’ll have to you soon
I’m writing a story today about the work promoted by the IPA on the EPBC Act and the methodology behind the claim that there has been a “445 per cent” increase in regulations under the act since 2000.
I need to file before the end of the day, but the story raises some strong criticisms of the RegData methodology the IPA used. I want to give the IPA the chance to respond.
- A UC Hastings law professor who has written a critique of RegData, says the method makes no attempt to measure the effectiveness of regulations. Nor does it differentiate the use of the keywords used in the RegData to see if they actually relate to a regulation, or to something spurious.
- That law professor, and an Australian environmental lawyer, say the motive for RegData is not to improve regulations, but is a “political project” developed at a right-wing US think tank (referring to Mercatus) that is ideologically driven.
The Guardian | Australia