This post is a more complete statement of my conclusions based on the analysis in Parts One and Two of this series. As I’ve explained in Part Two, there’s no reason in the Reinhart-Rogoff (R-R) data to believe that the debt-to-GDP ratio has a negative impact on growth. Ironically, that’s because their data set is terribly biased in its incompleteness, and was constructed to try to prove that there was a negative relationship between the debt-to-GDP ratio and economic growth. The interests supporting the RR work, both in its inception, and in disseminating its original results, were clearly trying to develop a basis for saying that since there is such a negative relationship, the right thing to do when the ratio gets too high (over 90%) is to implement a program of austerity aimed at deficit reduction, more or less drastic, depending on the individual case.
Of course, there may well be a relationship between debt-to-GDP ratios and economic growth in nations lacking non-convertible fiat currencies and floating exchange rates, and and/or having external debts in currencies they cannot issue. However, the R-R data set didn’t include those variables, so that analysis can’t be done without augmenting the data set. In such nations, MMT theory suggests that Government Budget Constraints (GBCs) on deficit spending, such as those we find in Eurozone nations would create a negative relationship between debt-to-GDP ratios and growth.
In fiat sovereign nations, such as the US, the UK, Australia, Japan, etc. we might also have the presence of an indirect relationship between variations in the debt-to-GDP ratio and economic growth through the actions of politicians who believe in austerity ideology pulling back on government deficit spending and consequently having a negative impact on economic growth through that mechanism. But to test for that self-fulfilling prophecy, and also for the negative relationship in nations subject to a GBC, someone will, again, have to augment the R-R data set and re-analyze it to include currency regime variables
In addition, we need to build on the biased and incomplete R-R data set to begin to test alternative hypotheses about the effects of austerity and different types of fiscal and monetary policy on different outcome variables and on feedback relationships from those outcome variables to economic growth and much more. When Matthew Berg and Brian Hartley say: “We suggest that simultaneous equations models may offer a way forward on the “frontier question” of causality,” they are also saying that other possible causes of both economic growth and debt-to-GDP ratios must be included in richer theories of economic dynamics, if we want to understand the place of both growth and debt in the broader context of what matters to people.
What matters to them are economic and social value gaps related to the idea of Public purpose like these:
— the gap between actual output and projected “full” output;
— High involuntary unemployment vs. full employment;
— Price stability vs. inflation or hyperinflation;
— Minimum wage vs. a living wage;
— No operative right to health care for everyone;
— social exclusion and the loss of personal freedom;
— skill deterioration due to unemployment;
— psychological harm such as sense of identity, self-respect, and sense of
— much greater ill health and reduced life expectancy than necessary;
— loss of motivation to live a full empowered life;
— deterioration of social relations, communities, social networks, and family life;
— increasing racial and gender inequality;
— increasing educational inequality;
— decreasing equality of opportunity;
— loss of social values and sense of individual responsibility;
— increasing economic inequality over time;
— increasing poverty;
— increasing crime rates including increasing use of control frauds by
important economic institutions;
— Failure to prosecute and punish people who commit control frauds;
— The collapse of real estate values and the destruction of the wealth of
working people after the crash of 2008;
— increasing anger against economic and political elites that get more and
more and more wealthy, and more and more immune to the rule of law;
— increasing political inequality undermining political, social, and economic democracy;
— increasing political unrest and threats of political violence both from the privileged and those seeking change.
— increasing environmental degradation;
— Increasing climate change/global warming.
— the gap between current energy foundations of our economy and new energy foundations based on renewables.
It will involve more of an effort to gather the necessary data in some of these areas than in others, and doing this kind of thing is a multi-year job. But it’s imperative that something like it gets done, because the kind of narrowly focused data set created by R-R is biased towards the concern of neoliberal ideology with debts, deficits, inflation, and economic growth, and its lack of concern with the impact of its favored economic policies on a range of outcomes important for most people. We need to be gathering data on those outcomes and analyzing the past, present, and likely impacts of alternative fiscal and monetary policies on them. In short, we need to be gathering data that allows us to test the impact of alternative fiscal policies on public purpose.
Finally, we must ask why there wasn’t a greater outcry from progressive activists and economists when the R-R study first appeared and they failed to make their data available for re-analysis and replication. After all, everyone who read their work and who knows even a little about quantitative analysis in the social sciences could see that it was based on a very superficial two-variable cross-country global data analysis, and that any result they reported had to present a false picture of causality.
This is true because you can’t provide a thorough analysis of causality between two cross-country variables without including additional variables and doing time series analysis at the national level to establish causal ordering and partial out spurious correlations. This has been well-known in the social sciences for at least 50 years.
MMT economist Randy Wray has called the R-R study “crap.” He’s right; for all the reasons just advanced, it was crap from the get-go. It presents an argument of partisan advocacy, not one of economic scientists making a conscientious effort to get at the truth.
So, the question is why has it been it challenged so little since 2010? It’s true that some economists provided critiques. But the discipline as a whole was respectful. The criticism was civil, when it should have expressed outrage. Everyone treated the critical exchanges as a matter of “he said, she said,” even though every economist who does any data analysis must have recognized the very simplistic level of R-R’s data analysis.
So, again I ask, why didn’t economists make ‘em prove it? And why did policy makers accept the findings so easily? You can’t tell me that the top economists in the Obama Administration, in the UK, and the Eurozone couldn’t see the nakedness of their co-emperors. They chose not to see.
I think there’s really no mystery here. Neoliberal elites wanted to believe in the austerity fairy tale for various reasons, including perhaps a desire to widen the wealth gap between the very rich and the middle class, and also a belief that belt-tightening in welfare states has moral value for the population subjected to that belt-tightening, though not for them, of course. For them, R-R was just window dressing for the financial sadism they wanted to implement anyway. If you doubt this characterization, then pay close attention to interviews of Erskine Bowles and Alan Simpson sometime. The are exhibit A.
But for the progressives, and others opposed to austerity, the R-R work should have immediately become a target of opportunity for educating the public about “junk” economic studies relied upon by politicians to justify their favored policies. Opposition to the study should have taken the form of telling people never to trust simplistic two variable analyses using cross-sectional rather than time series data to develop causal explanations. It should have taken the form of a demand for the economists and policy makers to prove what they say rather than just wave around a fig leaf that couldn’t possibly, and in the end did not, prove a thing about the desirability of austerity in modern economies.
But none of this occurred. And partly as result of this dog who never barked, millions around the world live with economic hardship lasting for years. Millions lost their homes. Millions went into bankruptcy, and many thousands needlessly died from lack of medical care and are still dying today.