Trends Worth Tracking

Given our usual way of doing economic analysis, the question of which institutions will most increase economic welfare rarely depends much on the exact values of the sorts of parameters social scientists and the media track with such enthusiasm and concern. (more)

I’ve complained before about useless trend tracking, but I don’t mean to suggest that all trends are uninteresting. Some trends tell us about how well our institutions are functioning. For example:

Accounting statements are getting less and less representative of what’s really going on inside of companies. … The finance industry showed a huge surge in the deviation … from 1981-82, coincident with two major deregulatory acts that sparked the beginnings of that other big mortgage debacle, the Savings and Loan Crisis. The deviation … reached a peak in 1988 and then decreased starting in 1993 at the tail end of the S&L fraud wave, not matching its 1988 level until … 2008.

Neither manufacturing nor IT showed the huge increase and decline of the deviation … that finance experienced in the 1980s and early 1990s, further validating the measure since neither industry experienced major fraud scandals during that period. The deviation for IT streaked up between 1998-2002 exactly during the dotcom bubble. (more; HT Tyler, Thoma)

Now that’s a trend to make me stand up and take notice! Similar parameters where I’d want to watch trends:

  • Marriage cheating and cuckoldry
  • Biased scientific papers, referee agreement
  • Wrongful convictions, faked evidence
  • Biased rulings by sports referees

These sort of trends track the health of specific institutions. When such an institution starts failing, we should be especially eager to reform it, using economic theory to suggest which reforms might be most effective.

GD Star Rating
Tagged as: , ,
Trackback URL:
  • Paul Crowley
    • Vladimir Nesov

      They work for me.

      • Paul Crowley

        They got fixed :-)

  • Douglas Knight

    That list is quite heterogeneous, a mix of objective and subjective, proxies and explicit cheating. I think the most promising item on the list is referee agreement because it is objective but probably not subject to fabrication as much as asking people whether they cheat. Though it is a very loose proxy and I’m not sure what I’d conclude from it. Do journals keep records of their referee reports? Or do they recognize them as a threat to the authority of the editors and destroy them?

  • Sister Y

    The Benford’s Law thing is a lucky match – accounting’s naive purpose (providing accurate information about transactions) happens to be its actual purpose, from the point of view of society (non-insiders); and there happens to be a way to measure whether data are generated by natural-ish processes or faked by insiders.

    The purposes of the institutions of marriage, science, criminal justice, and sports are much more varied – both in the sense that different stakeholders conceive of them as having different purposes, and in the sense that they may have multiple purposes to the same stakeholder (including purposes at odds with the naive purpose of the institution). Of the example domains, accuracy for its own sake (as with accounting) only seems to be a strong purpose within science; in all other cases, accuracy seems secondary, only valued to the degree it is instrumental to various human social purposes. For instance, paternity confidence is only one of many values humans may derive from marriage; paternity is nice and measurable, and a naive observer might attribute a rise in paternity confusion to a breakdown of the institution of marriage. However, the time-tested methods that prevent paternity confusion (like niddah and purdah) preclude what most people see as important purposes of marriage (not to mention human flourishing in general).

  • Sister Y

    Also – what is the probability that firms figure out the obvious fact that they need to fake Benford-consistent data sets in order to achieve their deceptive aims, and are able to do so at low cost?

    In general, tracking a variable as a proxy for deception will increasingly track lack of sophistication of relevant participants instead.

  • Mark M

    To summarize, we shouldn’t track useless trends and should track useful trends. Got it.

    What might a trend in cuckoldry tell you? How might that information be used? This is a real question – I really don’t know how you would qualify a trend as good or bad (is more cuckoldry optimal? or less?), or how knowing about the trend would be useful. Marketing? Social engineering?

  • Robin Hanson

    Douglas, few journals ever allow outsiders access to their referee reports.

    Sister Y, yes firms will learn to fake number distributions to avoid detection, but that will still let them fake overall return numbers, and so we can look at such overall numbers to see overall faking levels.

    Mark, higher cuckoldry levels say marriage is less succeeding as a commitment. We might then consider offering stronger commitment options.

  • Doc Merlin

    This all fits the Austrian story that a business cycle is an accumulation of mistakes.

  • Pingback: TheMoneyIllusion » Are corporate profits actually very high?