Saturday, November 24, 2018

On Mariana Mazzucato's The Value of Everything

Mariana Mazzucato's The Value of Everything: Making and Taking in the Global Economy is a popular book. She argues that we should change our ideas of what we consider productive and unproductive jobs and activities. She presents a brief overview of the history of economics focused on the classical and neoclassical theories of value, describes how Gross Domestic Product (GDP) is calculated from the System of National Accounts (SNA), examines how finance came to be mistakenly considered productive of value added in the SNA, and argues that government activities have been mistakenly considered as necessarily unproductive.

Value theory is a good way of organizing a popular history of economics, although it leaves out German historical schools and institutionalism. Mazzucato praises Petty for an early attempt at drawing up national accounts. She says he did not have a theory of value. (From secondary literature, I am under the impression he talked about labor being the father and land the mother of value.) She says Quesnay's Tableau is the "first spreadsheet". She has Adam Smith putting forth a labor theory of value. She notes that he had two theories of productive and unproductive labor: work that provides a physical product versus work for out of capital. (This is straight out of Marx's Theories of Surplus Value.) With Smith, Ricardo, Marx, she doesn't get into details of labor embodied, labor commanded, the transformation problem, and so on. She distinguishes between objective versus subjective value theories. Neoclassical theory is subjective. In some sense, it gets rid of value theory for a theory of price. Producing anything that people are willing to pay for becomes productive.

I wished she had included an illustrative table when she turns to national accounts. She notes that they are drawn up with many conventions, not all of which can be derived from neoclassical theory. She popularizes various objections to how the Gross Domestic Product (GDP) is calculated and used. Along with feminist economists, she notes that household production is not counted. Pollution and cleaning up are more productive of GDP than not polluting in the first place. She notes that certain government activities appear in GDP as productive, even though is not clear from neoclassical theory that (some?) government activity can be productive.

Mazzucato has lots to say about finance. In classical thought, finance would probably considered unproductive, with returns to finance akin to rents. But that is not currently the case. She mentions the idea of privatized Keynesianism, in which increased borrowing by those not well off maintains effective demand in an era in which ideology promotes austerity for government. She looks at how Milton Friedman and Michael Jensen promoted the maximizing of shareholder value, disregarding other stakeholders in corporate management.

Mazzucato argues a need exists to transition from looking at government as a provider of public goods to looking at how government can be productive of value. Health care and information technology are two industries that provide examples of this perspective. In both cases, government funds basic research, with private industries taking profits, including through patents and copyrights. In health care, the idea of Quality-adjusted Life Years (QLYs) is used to justify prices with extreme markups, that cover much more than private Research and Development costs. How many QLYs would be sacrificed if a certain drug was not there? In the view of institutionalism, going back to Veblen, technological innovation combines the activities of many, although the opposite view is embodied in current intellectual property law. From Mazzucato's view, government has been and can be productive.

My initial reaction to the overall thesis of this book was that it is too idealistic. The ruling ideas are the ideas of the ruling classes. Changing those ideas requires changing the material base. I am not sure I fully get her idea that how we draw the boundary between activities and jobs that are productive and unproductive is performative. Mazzucato gives examples of finance, of other activities that would be formerly classified as unproductive grabbing of rents, and of government activities. I came to agree that there is a need for the development and a public discussion of a conceptual frame of who produces what.

I think of this book as a popularization of certain practical and policy ideas drawing on heterodox economics. Mazzucato does not make a point of discussing or drawing boundaries between heterodox and orthodox, non-mainstream and mainstream economics. She draws on, say, Duncan Foley or Joseph Stiglitz when each is useful for her points. I like that this shows that heterodox economists are economists. She could have been more explicit, perhaps with loss of rhetorical efficacy, on when she draws on Marx.

Overall, I recommend this book. I wonder what Dean Baker or Robert Reich would make of this book. Anwar Shaikh and Ahmet Tonak would also have a reaction.


Blissex said...

Interesting summary, and thanks for the pointer to Anwar Shaikh and Ahmet Tonak's book, even if it is about "wellbeing" rather than "production".

My impression is that as to GDP there are three distinct issues:

#1 Given the "historically correct" definition of GDP, what should be counted as GDP?

#2 What is the usefulness of that definition?

#3 If that definition is not very useful, what should we use instead?

The "historically correct" definition of GDP as "Gross Domestic Production" was as a vector/list of physical quantities, and its purpose was as the "Production" indicates to gauge *roughly* how production is doing, because that is pretty fundamental, even if not the whole story, if only because of the omission of capital shrinkage/depreciation, and that answers #2: it is about production.

What is normally discussed as "GDP" is actually not "Gross Domestic Production", but it is the GDP index, obtained by multiplying the GDP vector with an arbitrary vector of prices, which itself can be the result of some additional calculations (e.g. hedonic adjustments).

But the GDP index is supposed to be the same numerically as GDI, a wholly different entity, and actually pretty much all the current discussions are about GDI, not about GDP, or even the GDP index.

As #1 the rule is that only *final* production of goods, services and capital counts as GDP, so intermediate goods, services and capital don't. Which for example means that most of education, research and finance should not be counted, because they are performed as an input to production, while military and government bureaucracy should be counted, because they are performed regardless of the production of anything else, that is they are final.

As to #3 the "historically correct" definition of GDP is highly useful within its limitations; most of the criticism against GDP is actually criticism of the GDP index, that is aggregate GDI, as only the direst neoclassicals believe that "income" measures exactrly contribution, while it has a large component of extractive power.

As to the limitations of GDP, the main one is that "Gross", that is not taking into account various forms of capital depletion. But while physical production is often more easily measured, measuring capital and its depletion is very difficult, and so the people who set up the national accounts took a shortcut.

Blissex said...

«What is normally discussed as "GDP" is actually not "Gross Domestic Production", but it is the GDP index»

Even so often what is actually discussed is not even the GDP index, but total GDI (Gross Domestic Income) which is quite a different thing; under some conditions the GDI index and total GDI should match, but the conditions are unrealistic.

Discussions on GDI (even if called GDP) are particularly beloved of neoclassical-leaning Economists, but they are very related to production (or productivity).

Blissex said...

obvious correction: "but they are not very related to production (or productivity)."

Blissex said...

Further very late correction: "under some conditions the GDI index and total GDI should match" should be "under some conditions the GD*P* index and total GDI should match"