What is new and essential in economics?

By "new," restrict yourselves to 1990 and afterwards.  I see mid-1980s as the end of a great era in economic theorizing.  Take game theory, principal-agent theory, and the economics of information, and apply them to everything, for better or worse.  This was an exciting, indeed intoxicating, time to learn economics.  While applications continue, we have run out of new ideas on those fronts.  Experimental economics is completely Nobel-worthy, but it is now over forty years old.  What are the next breakthroughs or the breakthroughs which have just been made?  Comments are open…

Comments

The new breakthrough will come from sociological economics and the links between formal and informal incentive mechanisms. Social networks explain certain economic outcomes and are not taken into account in present economic models. See Mark Granovetter homepage for some examples. http://www.stanford.edu/dept/soc/people/faculty/granovetter/granovet.html
In many of the realms that economics has entered in the last couples of decades, informal networks shape decision making as mas a formal incentives do. The effects of these linkages should be considered if we want to enhance our prediction capabilities.

Behavioral economics - especially the interaction between sociology and econ, the decreased tensions among their practioners when communicating with each other. Also, the experimental econ that requires technology that didn't exist until very recently (fMRI's and such). Perhaps these two fields will jointly influence micro theory on taste formation, which might revolutionize micro beyond the mere Thaler type insights.

The application of economic modeling tools to the things that psychologists have figured out about why people believe what they do and how they can be persuaded to change their beliefs. People are already thinking about this in "standard" econonmics terms (such as thinking about the effect of media ownership restrictions on persuasion). I think it will branch out into political economy stuff. Nothing is more important than what people believe.

All those Nobel prize winning insights from the late seventies and early eighties turned out to be exceptions, not the rule, and so you have economists like Stiglitz and Krugman arguing for every type of intrusive regulation imaginable because they 'proved' that markets are inefficient. I'm not sure that's such a great era.

What is a consensus among economists that was not there in 1989? That socialism is inferior to market economies in producing wealth and taking care of the environment. Of course, this was mainly an empirical matter, but it became too obvious to argue, and Samuelson did for decades, that they were comparable in growth, and better at using their resources, etc.

I second David J. Balan's point from above: Nothing is more important than what people believe. Research on heterogeneous priors, belief formation, how mistaken or biased beliefs get exploited in the market place, and so on is imho the most exciting new development in mircoeconomics.

Another extremely important development was the spread of "incomplete contracts" - thinking across pretty much all subfields of applied micro and even into macroeconomics and business cycle theory. Pretty much all the important papers of the 1990s in labor, development economics and even international trade are based in some form on Hart and Moore's work.

Slightly off-topic, but probably the exciting new development we have all been waiting for: money economics!

http://www.journals.uchicago.edu/JPE/journal/issues/v114n3/31428/brief/31428.abstract.html

I should add that I am not trying to make fun of the paper (which I have not read yet). In fact I had the pleasure to interact a little with Keith Chen when he was a grad student at Harvard and have the utmost respect for him.

I think behavioral economics and nueroeconomics, basically the synthesis of psychology into economics to create more accurate "mono-economic" models and new theory resultant from the aggregation of these models.

sammler, above, is mistaken in his comment that game theory hasn't been applied
to the economics of the state. In fact it has, by Joseph Kalt and others, who
have elaborated a public goods theory of the state.
My favorite candidate for a Nobel prize is Benoit Mandelbrot, who has done
some good empirical work, such as a study of cotton prices, which overturned
Samuelson's view of how stock prices fluctuate. Mandelbrot also applied the
theory of fractals to financial markets.

My vote is for welfare economics and the foundations of utility theory. Check out my comment over at EconLog from a while back on "Prisoners of the Food Metaphor: Why Economists Misunderstand the World."

I see public choice theory and welfare theory as pushing in opposite directions. Individualist understandings of utility are inadequate because almost all enjoyment, other than of food-- enjoyment, not consumption; the term "consumption" is part of the reason that economists are prisoners of the food metaphor-- is inherently communal, which implies that coordination problems and micro-externalities are a more important constraint on human welfare than resource constraints (especially as the world gets richer, materially).

This is related, a bit, to formal and informal incentive mechanisms, and I agree with the first commenter. But the general insight is that a lot of the things that people value can't be bought for money directly, yet they can and must be analyzed by economists if our understanding of human happiness is to be more than a travesty.

If welfare economics will (should) push microeconomics in a communitarian direction, macroeconomics should move in a libertarian direction, pushed partly by Austrian and real business cycle theories, but especially because of public choice theory, which shows how misguided is the economist's traditional posture of advisor to the government. The entity that economists have imagined they are advising-- a benevolent government trying to maximize the public welfare-- simply does not exist. A tension between economics and democratic politics may develop, on issues like immigration for example... which may also force a re-examination of the link between economics and ethics, as economists may have to entertain the question of whether the state's claim to be able to restrict beneficent labor flows as part of its "sovereignty" should deserve special deference, relative to other forms of violence.

I think the consensus emerging is that neoclassical methodology has exhausted its usefulness (and some might even agree its welcome). I expect heterodox thought finally becoming a little more mainstream. Experimental and Behavioral are just the beginning. Economics will finally get over themselves and welcome more insight from sociology (a relationship akin to that between physics and chemistry -- physics, however, really is foundational!). Reductionism is out and complex system theory is in. This revolution has barely taken root in the dismal science, whereas int physics, biology, climate science, and ecology it has. What is needed is a Samuelson of sorts, well-versed in the canonical mathematics and also in complex system theory (modern dynamics), to write a phenomenal new textbook with which to train the new generation of economists.

Behavioral and neuro are possibilities, but I still have to be convinced they can be stated in a rigourous enough way to make them more than just a collection of facts. Some of the suggestions above that seem the most likely are field experimentation, belief formation, and especially spacial economics, which at this point is almost practice without a theory.

Experimental economics might be 40 years old, but field experimentation is extremely new and has proven many of the prior conclusions of laboratory experimental economics wrong (and neoclassical price theory correct).

Alec, I am curious to know exactly what field experimentation has proven incorrect relative to laboratory experiments. Thanks.

AZ:

A couple of things that come to mind are:

John List's work on the endowment effect (indifference curves probably don't cross)

List's work on market anomolies on individual's with experience (not as many anomolies)

List's work on gifts exchange and labor output

Glenn Harrison's work on discount rates (not nearly as hyperbolic as people thought)

There are more, but note that field experimentation is an extremely new field of experimental economics (it was first used awhile ago, but it started being used as a method seperate from lab experiments only a little while ago).

End of the 90s technological changes enable cheap worldwide distribution of large data sets in electronic format at zero cost via the internet. Many sciences took immediately advantage of it to foster progress (the technology in question was developped for physics :), economists still haven't caught up with that trend, data is still hidden as a secret treasure not to be shared in this field.

Once we start to see big data sets available for all and economists start to use them to calibrate computer simulated agent models, a few good thing should happen to economics.

Alec:

I agree. Field experiments are a natural, and I think very promising, extension to lab experiments.
I just disagreed that field evidence already falsified results from the lab. I think that there is just not enough evidence to conclude that. But the discussion initiated by List´s experiments is definitely exciting and I am looking forward to what comes out of it.

AZ,

No time right now for a detailed response, but the crucial difference between a field experiment and collecting data from a business is the ability to randomize a treatment. Said differently, when you collect data from a business then pretty much every single variable you observe is completely endogenous - the contracts employees or customers are being offered or choose, the assignment of employees to tasks, the targeting of marketing campaigns, and so on. This makes it virtually impossible to ever measure true causal effects of any variable. Field experiments, wherever feasible (and that's not often), allow you to overcome this very fundamental problem of standard empirical research.

Quantative Macro Theory. It's the macroeconomist's answer to Deirdre McCloskey's question, "How big is big?"

Selected items (not all of which fit into the QMT box, but close enough for government work):

-Job Creation and Destruction (students love the metaphor, for good reason)

-The revival of growth theory--which brings in lots of data to see which theories are more plausible, and which are quantatively important.

-Industry-level and firm-level productivity studies.

-The quantitative wave of trade/urban economics research: A new attempt at mesoeconomics---to which Lucas is an important contributor.

In sum: Basically, anything with a production function and some kind of calibration is going to grab the attention of a lot of young economists these days....for good reason. We've moved beyond t-stats and into the realm of "What is really important in explaning large parts of the world around us?"

Q: Where do I learn more about this exciting line of research?

A: By checking out the Society for Economic Dynamics's website. Their annual meeting always has lots of interesting QMT research...Just flip through a few abstracts to get the main idea.....

http://www.economicdynamics.org/society.htm

AZ,

Commenterlein is completely right about the key distinction between naturally occuring data and data obtained from a field experiment.

As for this point about experience: The field is by no means the used primarily because people are more experience. In fact, List runs a baseball card experiment with little children to see how varying experience can effect price theory (also you can gain experience in the lab, it just takes awhile).

Also, you are right that one area of field experimentation is just running lab experiments with normal people (as opposed to college students). I think these are called artefactual field experiments, but I could be wrong. Then there are other varying degrees of field experimentation which more closely approaches real life action. Of course running these, in some situations, can be more expensive than the lab, but if in doing so, you can overcome a lot of the obstacles inherent in lab experimentation (namely experimenter demand effects, framing effects, etc.) then you have really achieved something. Check out List's working paper on the lab cowritten with Levitt that deals with this topic. In that, the conclusion is essentially that lab experiments are great to test new mechanisms, but in no way should the results produced there be extrapolated to the outside world until those same mechanisms have been rigorously tested in the field.

Also, the whole MIT poverty lab stuff is an example of field experiments, so you can get cool theory testing with this stuff and rigorous ways of testing the effects of various mechanisms that could aleviate poverty.

This conversation on List's field experiments is interesting and I have just reviewed some of the studies cited above. I agree that field experiments are much different than lab experiments and that they are essential.

On a point of substance, JL, your sample size comment does not seem correct on his econometrica paper. He finds gift exchange in the first few hours of both exercises and this gift exchange result disappears with time. This has nothing to do with sample sizes since he finds gift exchange and than rejects it later.

More broadly, this result is intriguing because it says that short run elasiticities measured in the lab might not be good estimates of the long run elasticities that we are typically interested in the field. I take this as the big point of that paper and something that I had never thought about.

Some thoughts:

1) There is a lot of exciting stuff happening in regard to micro-econometrics and individual heterogeneity. This area is the subject of Heckman's nobel lecture published in the Journal of Political Economy. For example, the interpretation of instrumental variables estimators has changed dramatically in the last few years, although this change has yet to spread very far beyond the more teched-up parts of labor economics. I am rather surprised that no one else has mentioned this.

2) Networks are already hot in economics; the trouble with them is that, as Manski has shown, it is very hard to make empirically based causal claims about things like peer effects. More generally, sociology is full of excitement about the more serious application of empirical methods drawn from economics and statistics. My sense is that even many sociologists are tired of endless rants about the holy trinity of race, class and gender dressed up as research.

3) The economics of crime is very hot. I just attended an excellent conference that very profitably brought together economists and criminologists. Given the astounding fraction of the US population in prison or otherwise enmeshed with the criminal justice system, this is important stuff.

4) I agree about behavioral economics, though it is producing a lot of bad theory. Little insight is produced by generating new sets of axioms consistent with the latest insight on the behavior of undergraduates in laboratory experiments. I think the big potential here is empirical, guided by applied theory that views cognition as the solution to a problem regarding the allocation of scarce mental resources.

5) I agree about spatial/urban economics. Ed ("from hell" as he was called by the students he TA'd at Chicago) Glaeser has almost single handedly revived this field.

6) I agree about policy-oriented field experiments in labor and development though I would note that these were not invented at the poverty lab at MIT. In fact, social experiments have been widely used in the US for the last twenty years for the evaluation of welfare to work and employment and training programs.

7) I am less sanguine about neuroeconomics at the moment. Looking at MRIs is good cocktail party fun but I do not see that we are learning much yet. Still, in the spirit of letting 1000 flowers bloom, I think it is worth pursuing.

Jeff Smith

evolutionary economics. This will be the future.

Economics is ripe for a Kuhnian scientific revolution. I agree with those who have highlighted the importance of behaviroal economics, though I would add what is *really* the relevant research here is that which seeks to reunify economics with behavioral sciences more broadly. So some other social scientists dont know a lot of math. Big deal. They still have a lot to teach us economists. In this area, Samuel Bowles and his cohort are on the cutting edge. Check out http://www-unix.oit.umass.edu/~bowles/. Evolutionary economics--particularly the co-evolution of institutions and individuals is a branch of this research agenda, and Bowles' textbook is a great introduction to the subject. In contrast to the 1980s heyday, what we are looking at now is specifying the **macrofoundations of microbehavior**.

Here, the role of institutions is central (a point which has been known since the classical political economists), but which has never really been dealt with adequately in theoretical formalization or in empirical applications. Cooking up some index of economic freedom (like Heritage Foundation or Fraser Institute do) or subjective measures of governance quality (like the World Bank does) and dumping them in the right hand side of an OLS specification are not shedding light on the matter.

Behavioural economics, definitely. Frey, for example, has done his most substantial work since 1990, and it is likely to become more relevant over time.

Non-monetary inflation can be stopped.

"People today use the term `inflation' to refer to the phenomenon that is an inevitable consequence of inflation, that is the tendency of all prices and wage rates to rise." Ludwig von Mises - "Inflation: An Unworkable Fiscal Policy".

All prices do not rise. Only the prices of variable real value non-monetary items while many constant real value non-monetary items are not fully updated and many are not updated at all.

The second inevitable consequence of inflation is the tendency of many constant real value non-monetary items NOT to rise at all - during the Historical Cost era while some constant real value non-monetary items are not fully updated.

Inflation today has and always had a second consequence during the 700 year old Historical Cost era.

Inflation has a monetary consequence, called cash inflation refered to above by Ludwig von Mises and defined as the economic process that results in the destruction of real economic value in depreciating money and depreciating monetary values over time as indicated by the change in the Consumer Price Index.

Inflation´s second consequence is a non-monetary consequence defined as Historical Cost Accounting inflation which is always and everywhere the destruction of real economic value in constant real value non-monetary items not fully or never updated (increased) over time due to the use of the Historical Cost Accounting model or any other accounting model which does not allow the continuous updating (increasing) in constant real value non-monetary items in an economy subject to cash inflation.

Inflation´s second consequence is solely caused by the global stable measuring unit assumption.

The stable measuring unit assumption means that we regard the annual destruction of a portion of the real value of our monetary unit by cash inflation in low inflation economies as of not sufficient importance to update the real values of constant real value non-monetary items in our financial statements.

This results in the destruction of at least $31bn in the real value of Dow companies´ Retained Income balances each and every year. Globally this value probably reaches in excess of $200bn per annum for the real value thus destroyed in all companies´ Retained Income balances.

The International Accounting Standards Board recognizes two economic items:

1) Monetary items: money held and "items to be received or paid in money" – in terms of the IASB definition.

2) Non-monetary items: All items that are not monetary items.

Non-monetary items include variable real value non-monetary items valued, for example, at fair value, market value, present value, net realizable value or recoverable value.

Historical Cost items valued at cost in terms of the stable measuring unit assumption are also included in non-monetary items. This makes these HC items, unfortunately, equal to monetary items in the case of companies´ Retained Income balances and the issued share capital values of companies without well located and well maintained land and/or buildings or without other variable real value non-monetary items able to be revalued at least equal to the original real value of each contribution of issued share capital.

The stable measuring unit assumption thus allows the IASB and the Financial Accounting Standards Board to conveniently side-step the split between variable and constant real value non-monetary items. This is a very costly mistake in low cash inflation economies - or 99.9% of the world economy.

Retained Income is a constant real value non-monetary item, but, it has been in the past and is, for now, valued at Historical Cost which makes it, very logically, subject to the destruction of its real value by cash inflation in low inflation economies - just like in cash.

It is an undeniable fact that the functional currency's internal real value is constantly being destroyed by cash inflation in the case of low inflation economies, but this is considered as of not sufficient importance to adjust the real values of constant real value non-monetary items in the financial statements – the universal stable measuring unit assumption which is the cornerstone of the Historical Cost Accounting model.

The combination of the implementation of the stable measuring unit assumption and low inflation is thus indirectly responsible for the destruction of the real value of Retained Income equal to the annual average value of Retained Income times the average annual rate of inflation. This value is easy to calculate in the case of each and very company in the world with Retained Income for any given period.

Everybody agrees that the destruction of the internal real value of the monetary unit of account is a very important matter and that cash inflation thus destroys the real value of all variable real value non-monetary items when they are not valued at fair value, market value, present value, net realizable value or recoverable value.

But, everybody suddenly agrees, in the same breath, that for the purpose of valuing Retained Income – a constant real value non-monetary item – the change in the real value of money is regarded as of not sufficient importance to update the real value of Retained Income in the financial statements. Everybody suddenly then agrees to destroy hundreds of billions of Dollars in real value in all companies´ Retained Income balances all around the world.

Yes, inflation is very important! All central banks and thousands of economists and commentators spend huge amounts of time on the matter. Thousands of books are available on the matter. Financial newspapers and economics journals devote thousands of columns to the discussion of the fight against inflation.

But, when it comes to constant real value non-monetary items:

No sir, inflation is not important! We happily destroy hundreds of billions of Dollars in Retained Income real value year after year after year.

However, when you are operating in an economy with hyperinflation, then we all agree that, yes sir, you have to update everything in terms of International Accounting Standard IAS 29 Financial Reporting in Hyperinflationary Economies: Variable and constant real value non-monetary items.

But ONLY as long as your annual inflation rate has been 26% for three years in a row adding up to 100% - the rate required for the implementation of IAS 29. Once you are not in hyperinflation anymore (for example, Turkey from 2005 onwards), then, with an annual inflation rate anywhere from 2% to 20% for as many years as you want, you are prohibited from updating constant real value non-monetary items. Then you are forced by the FASB´s US GAAP and the IASB´s International Accounting Standards and International Financial Reporting Standards to destroy their value again – at 2% to 20% per annum - as applicable!

For example:

Shareholder value permanently destroyed by the implementation of the Historical Cost Accounting model in Exxon Mobil’s accounting of their Retained Income during 2005 exceeded $4.7bn for the first time. This compares to the $4.5bn shareholder real value permanently destroyed in 2004 in this manner. (Dec 2005 values).

The application by BP, the global energy and petrochemical company, of the stable measuring unit assumption in the accounting of their Retained Income resulted in the destruction of at least $1.3bn of shareholder value during 2005. (Dec 2005 values).

Royal Dutch Shell Plc, a global group of energy and petrochemical companies, permanently destroyed $2.974 billion of shareholder value during 2005 as a result of their implementation of the stable measuring unit assumption in the valuation of their Retained Income. (Dec 2005 values).

Revoking the stable measuring unit assumption is actually allowed this very moment by IAS 29 but ONLY for companies in hyperinflationary economies. At 26% per annum for three years in a row, yes! At any lower rate, no!

It is prohibited by US GAAP and IASB International Standards for companies that are operating in a low inflation economy.

That means the following at this very moment in time: Today all companies in, most probably, only Zimbabwe (1000% inflation) are allowed to update all their variable real value non-monetary items as well as all their constant real value non-monetary items.

But not the rest of the world.

The rest of the world is forced by current US GAAP and IASB International Standards to destroy their/our Retained Income balances each and every year at the rate of inflation because of the implementation of the stable measuring unit assumption whereby we are all forced to regard the change in the value of the unit of account - our low inflation currencies - as of not sufficient importance to update the real values of constant real value non-monetary items in our financial statements.

We are forced to destroy them year after year at the rate of inflation till they will reach zero real value as in the case of Retained Income and the issued share capital values of all companies with no well located and well maintained land and/or buildings at least equal to the original real value of each contribution of issued share capital.

The 30 Dow companies destroy at least $31bn annually in the real value of their Retained Income balances as a result of the implementation of the stable measuring unit assumption. Every single year.

Retained Income can be paid out to shareholders as dividens. Poor Dow company shareholders. They will never see that $31bn of dividens destroyed each and every year.

We have all been doing this for the last 700 years: from around the year 1300 when the double entry accounting model was perfected in Venice.

When we do this at the rate of 2% inflation ("price stability" as per the European Central Bank and as per Mr Trichet, the president of the ECB) we are forced to destroy 51% of the real value of the Retained Income balances in all companies operating in the European Monetary Union over the next 35 years - when that Retained Income remains in the companies for the 35 years - all else except cash inflation being equal.

Each and every one of those 35 years will be classified as a year of "price stability" by the ECB and Mr Trichet. Mr Trichet will not be the president of the ECB in 35 years time.

I think we will do ourselves a great favour by revoking the stable measuring unit assumption as soon as possible.

FREE DOWNLOAD : You can download the book "RealValueAccounting.Com - The next step in our fundamental model of accounting." on the Social Science Research Network (SSRN) at http://ssrn.com/abstract=946775

--------------------

Nicolaas J Smith
http://www.realvalueaccounting.com/

Relatively new and essential topic: understanding (and accepting) the importance of randomness as a phenomenon, an accurate description, a confounding effect in policy attempts and both limit and challenge in human endeavour.

Best commencing summary post 1990 Taleb's "Fooled by Randomness" 2002. Proof of lag effects and populism Taleb's "Black Swan" 2007. Hidden (some implicit) historical revelations (20th century only) Kendall, Arrow, Markowitz, Fama, Jensen, Sharpe, Black.

I agree with the first commenter. But the general insight is that a lot of the things that people value can't be bought for money directly, yet they can and must be analyzed by economists if our understanding of human happiness is to be more than a travesty.

Comments for this post are closed