Thursday, February 20, 2014

On double-blind and triple-blind refereeing in journals.

Triple-blind refereeing would have been a great idea forty years ago. I am not sure it will work today for the following reason.

Some feedback on one's research well before the paper is sent to the journal is very important to most researchers. Most authors therefore publicise the paper in limited circles, say one's friends or a small social network of   researchers in the same area. Over time, this social network evolves around one or two very influential people (usually either politically active in the area, well connected, or editors/editorial board members). so, irrespective of the double- or triple- blindness of the review the review in reality tends to be far less than blind. Also, by the citations in the paper under review it is not very difficult for the reviewers to guess a very small set of people from which the author(s) must come.

I learnt the above very early in my career (I may have mentioned this in earlier posts). My first year out of PhD program I wrote a paper and included it in the school's list of working papers. I did not know that the school sent copies of working papers to all AACSB schools. A few months later I was horrified to find a citation to the working paper in one of the most prestigious journals in the field. Soon thereafter, I received a message from a wellknown academic in the area that the paper had good ideas and "let's work on it". I did not respond because the paper was essentially complete. A few months later, I sent the paper to another prestigious accounting journal for publication. My bad luck, the paper went to the same lets-work on-it academic. The review essentially said that idea was stale because the working paper (which included results in theorems/lemmata as well as their proofs) was already cited (the other review was a conditional acceptance subject to revision). I was, however, lucky in that the journal editor was open-minded enough to send it to a third reviewer, to break the tie, who accepted it also conditional on minor editorial changes. To date, it is one of my papers that has been cited in very diverse fields (Management Science, accounting, Operations Research, Finance, Space Research, Economics, Game theory,...), but hardly in accounting, even though the topic had to do with accounting.  Most fascinatingly, it was cited by some operations researchers who I do not know but were heroes of mine when I was a doctoral student in Operations Research.

Researchers in lesser known schools will ALWAYS be at a disadvantage because they are usually not in the tightly-knit small social networks whose membership is usually by invitation. This perpetuates the inequalities. These social networks are our equivalent of the one-percenters.  The rest of us peasants are the 47 percenters in their opinion. THE POLE OF ACADEMIC RESEARCH, ESPECIALLY IN ACCOUNTING, IS VERY GREASY, and those who reach the top are usually those who hang on to the coat-tails of those little social networks.

I do not mean the above in a disparaging way. Such networks perform a very important function: mentoring of recent entrants in the field.


I  always tell my doctoral students that the trick is to get into those small social networks as early as they can. Hard as it is to get into those networks, persistence usually pays off. I realise that this is a very cynical view of the world, but then I hate to have my students suffer.

One can hope that the growth of the internet  and the avenues for dissemination of research PRIOR to submission (such as SSRN, citeseer, arxiv.org,...) will provide some solace to the rest of us. On the other hand, much of such good research will get cited once or twice and then forgotten because of the pressures to cite people central to the area of research. The academic equivalent of rich-getting-richer. (In this whole post must apologise to the French Sociologist Vilfredo Pareto, the inventor of power law in the context of income distribution,  for stating what is clearly obvious; the world is mostly an implementation of power laws).

Wednesday, February 19, 2014

On Accounting and Finance.

It is the better developed theory that, in my humble opinion, differentiates Finance from accounting.

 When I started my doctoral program way back in the early seventies, finance was in its infancy. I remember taking my first doctoral finance seminar where the textbook was "The Theory of Finance" by Fama and Miller. The first part of the book (which is essentially elementary micro-economics with the time variable thrown in) was very beautifully organised, but the second part  where stochasticity in finance is added in, was vague, far too tentative, and thoroughly disorganised; the book essentially fell apart towards the end. At least it felt that way to me since I had arrived with a whole load of courses in probability theory and stochastic processes.  Finance theory then was just being refined, and Fama/Miller text had made a premature entrance. (I think Fama would have done better if only he had listened more carefully to his illustrious dissertation adviser, the late Benoit Mandelbrot).

Then for twenty years I had virtually no interaction with Finance until I was asked to teach a graduate course in Accounting Theory. A Finance colleague asked me to take a look at Luenberger's book on Investment Science (as the title suggests, it deals with only one aspect of finance). I found the book fascinating. Somewhere along the way, Finance had found itself. That can not happen without good theory.

Unfortunately, accounting, in my opinion, has yet to find its moment of epiphany. Until then, we all will be plodding along, pretending that we have a good understanding of our domain.