Syndicate content

Should a one-time entry cost matter?

As I mentioned in my previous post, some critics have argued that too much attention is paid to the costs of starting a business. One way of stating the critique is that a cost that has to be paid only once ought not to have a big effect on a firm’s decisions. Arvind Panagariya of Columbia University makes this argument in his opinion piece for the Economic Times. He states:

…even within the narrow confines of regulatory regime, some policy dimensions emphasised by the index are of questionable value. For instance, the index attaches great importance to the costs incurred and time taken in starting a business. But when firms are entering a market with a horizon of several decades, does it matter whether it costs $500 rather than $5,000 and takes 20 rather than 200 days to start the business?

My question is whether or not the time horizon for a potential start-up firm is really one of several decades. Eric Bartelsman, John Haltiwanger and Stefano Scarpetta study firm survival rates (and a lot more) in this article. Figure 4 gives the probability that a new firm survives two, four, and seven years for several countries.

Survival rates

On one end of the spectrum is Mexico, where a new manufacturing firm only has about a 30% chance of surviving seven years. At the other end of the spectrum is Slovenia, where a new manufacturing firm has about a 70% chance of surviving seven years.

Given the observed time horizons for new firms, at least in some countries, the monetary and time costs of initial registration may be quite important.


Submitted by Ryan Hahn on
David, Isn't there an additional endogeneity problem here that would actually make the case for being concerned about registration costs even stronger? Presumably, only those firms that think they have a reasonable chance of lasting at least a few years would consider starting up and registering. There may be a whole range of ventures that are not even tried and are thus not captured in this data. If development requires "creative destruction" and a high churn rate of business creation and destruction, then intial costs could be very important.

Submitted by David Kaplan on
Ryan: Yes. One would assume that firms on the margin between being formal or not would have even lower probabilities of surviving seven years. I’ll have more to say about that in a future post.

May I add another aspect: firm size. When I read Arvind Panagariya's article last year I send him a comment as I think his critique of the Doing Business (DB) is right but as DB focuses on domestic SMEs and not on large domestic or foreign companies one has to consider the following: True, if firms are entering a market with a horizon of several decades - e.g. if the venture is based on an innovative product / business idea - it does not really matter if starting this business will cost $500 or $5000. However, for small and micro firms in traditional sectors the relevance of these costs will probably increase. So I think it is reasonable to assume that entry costs become more relevant with a decrease in firm size, isn't it? From my point of view, however, the more interesting question raised by this post is why new manufacturing firms in different countries have such different survival rates. Probably this depends on productivity, but then: do entry costs really matter? Firms with high productivity will register anyway as they can afford it. Firms with low productivity might not register if entry costs are x - so what is gained if we lower them to y so that these firms register? Then they are formal, but is this change in informal/formal status important for the regulator? I don’t think so, because the regulator’s/the government’s purpose is to raise tax revenue, which is only generated by firms with high productivity, not by firms with low productivity – no matter if they are formal or informal. Does this reasoning make sense?

Add new comment