This post is a review of the 2017 edition of the OECD’s annual publication Government at a Glance. I begin with a personal disclosure. The team leader for this publication is Zsuzsanna Lonti. I had the honor to be an adviser for her excellent doctoral dissertation on innovative workplace practices in Canadian government. We have stayed in touch in the intervening 25 years. She asked me only to have a look at the most recent edition of G@G, as it is referred to informally, and here are my observations.

In its essence, G@G consists of a host of governmental performance league tables for the 35 OECD member countries, four that are in the process of joining, and five other major economies (Brazil, PRC, India, Indonesia, and South Africa). The topics covered include public finance, public employment and pay, integrity of public institutions, budgeting processes and institutions, regulatory governance, public procurement, open government, innovation and digital government, risk management, results such as trust in government and public sector efficiency, and the quality of government services. Though the data for some of these league tables are taken from other sources, such as national income accounts, the OECD conducts its own surveys in many instances.

When an area becomes a focus of discussion among practitioners and scholars, the OECD is quick to launch a survey to measure it. This year’s report added public sector innovation, risk management, and the application of behavioral insights. The latter has turned out to be very topical in light the 2017 Nobel Prize for Economic Science being awarded to behavioral economist Richard Thaler. The report also introduced national rankings for measures of the quality of citizen service in education, health, and the justice system. Rankings –especially low ones – capture attention in ways that index scores do not.

As an innovation scholar, I was very pleased to see a survey that asked about the implementation of innovation labs, innovation funds, innovation awards, and a variety of human resource development structures identified in the literature as being supportive of innovation.

The standard table in G@G starts with two dimensions, the nations reporting and the bars for the indicator reported by each nation. Tables can become more complicated if the indicator has several components, which is usually handled by showing several color components to the bars. Comparisons over time are often handled by superimposing small circles, diamonds, squares, or triangles on the bars. It sometimes takes concentration to make sense of the tables. Personal disclosure: I am color-blind (red-green, red-black, grey-blue) so some of the visual distinctions escape me.

The tables do not indicate which OECD members didn’t provide the data or complete the survey for a given measure. As a form of suasion, I would encourage the G@G team to list the nations that are “NA” in any table.

It is important that the OECD’s tables track the latest major developments and show them clearly. Canadian Prime Minister Trudeau introduced gender balance in his cabinet, and that showed up clearly in the table comparing the percentage of federal ministers who are women in 2015 (under his predecessor Stephen Harper) and in 2017 under Trudeau (p. 99).

The Trump Administration has made it clear that it will pay little heed to many of the “good government/good governance” practices that G@G tracks. This raises the question of whether the 2019 and subsequent G@G’s will show the impact of Trump’s approach on the US’s performance. Consider stakeholder engagement in the regulatory process. EPA Administrator Scott Pruitt has made it clear that his door is always open to the representatives of business but as for environmentalists, not so much. Will the effect of this asymmetry be captured in future editions of G@G?

I hope that the next G@G takes a detailed and searching look at the impact of the Trump Administration. Trump still has not appointed an ambassador to the OECD by the way; perhaps if the OECD is too pointed in its criticism, he will withdraw American membership.

How is G@G used? Among practitioners, I would think that poor scores are an incentive to a nation to improve. This supports my suggestion for calling out the NAs, because an NA is likely indicative of poor performance.

Civil society groups can make use of the G@G league tables to pressure governments to improve their performance. If it doesn’t do so already, the OECD should proactively be in touch with civil society groups to make sure they are aware of its work. How strong, for example, is the relationship between the OECD and the Open Society Foundation?

I hope public administration scholars, political scientists, and other academics concerned about public sector performance are taking advantage of these tables. The number of cross-sectional observations (a maximum of 45) in any given year is a bit small for causal modelling, but a statistical study could use observations from several years to expand the number of degrees of freedom. Even if a research design doesn’t involve statistical modelling, I would still encourage the researchers to look at the OECD’s work on the variables they are interested in and contact the OECD to get the information that serves as background for those league tables.

I congratulate Zsuzsanna and her colleagues on all their efforts to gather data from all over the world to tell us what governments are doing to improve public management and deliver better public policies. It’s important work and they should be proud of it.