Regulatory Barriers and Entry in Developing Economies* John Bennett, Brunel University Saul Estrin, London School of Economics$ September 28, 2007 Abstract We model the effects of two common entry barriers - licence fees and bureaucratic delay - on the entry of firms into an industry that is new to a given developing economy. We find that the effects vary significantly across barriers. A licence fee alone reduces both the number of first movers and the steady-state number of firms, but in combination with bureaucratic delay can have the opposite effects. This is because, although the delay between application and receipt of a licence is assumed the same for all firms, delay changes the balance of advantage between first and later movers. Keywords: Entry, Entry Barriers, Developing Economy JEL Classification: L50, 014 * We thank the Department for International Development for supporting this research under DFID/ESCOR project number R7844. Earlier versions of the paper were presented at the European Bank for Reconstruction and Development, London, February 2005; the Latin American and Caribbean Economic Association Annual Congress, Paris, October, 2005; and the ASSA meeting, Boston, January 2006. We are grateful to the participants, and in particular, Dan Berkowitz and Hadi Esfahani, for their comments. We also gratefully acknowledge comments and suggestions from Robin Burgess, Ravi Kanbur, Daniel Kaufmann, Mark Roberts, Stefano Scarpetta and Kathy Terrell. The usual disclaimer applies. $ Corresponding author: Management Department, London School of Economics, Houghton Street, London WC2A 2AE; tel. 020 7955 6605; email s.estrin@lse.ac.uk

1 Introduction

A substantial literature exists that analyses new firm entry in developed economies

(see, e.g., Jovanovic, 1982; Evans, 1987; Dunn et al., 1989; Geroski, 1995, Eric-

son and Pakes, 1995, and Caves, 1998). Recently, research has also addressed the

issue of entry in developing countries, considering, mainly from the empirical per-

spective, the impact on the entry process of the comparatively weak institutional

framework typically found in these economies (see, e.g., Tybout, 2000; Roberts

and Tybout, 1996; Djankov et al., 2002; Klapper et al, 2006). In this paper, we

develop a simple model to consider the way that specific institutional features of

developing economies — licence fees for entry, and bureaucratic delays — influence

the scale and pattern of entry.1

At least since de Soto (1990), it has been suggested that high barriers to en-

try of new firms have led to relatively low entry in developing economies, with

damaging consequences for productivity growth (see, e.g., Roberts and Tybout,

1996). Bureaucratic constraints and corruption have also been argued to lead to

the emergence of the large informal sector often found in developing economies (de

Paula and Scheinkman, 2006), though some analysts of Latin America character-

ize the informal sector as entrepreneurial (Maloney, 2004). Djankov et al. (2002)

provide comparative data on 85 countries and find the average level of the regu-

latory barriers to be high in most developing economies. The barriers measured

1

include the number of procedures required to start a firm (which, between coun-

tries, varies from 2 to 21), the minimum time for start up (from 2 to 152 days), and

the official cost (from 0.5% to 460% of per capita GDP). Significantly, they find

that countries with heavier entry regulations have larger unofficial economies and

higher levels of corruption, though they do not explore the relationship between

entry rates and regulations. However, Scarpetta et al. (2002) find that the rate

of entry of small and medium enterprises is negatively related to the number of

regulations, especially those in the product and labour markets (see also Desai et

al., 2003). The entry regulations is investigated in a huge cross-country enterprise

sample by Klapper et al. (2004). They find that, for industries with high entry

rates in the US, relative entry is disproportionately lower in countries in which

regulation imposes greater costs. They conclude that entry regulations do hinder

entry, especially in industries that ‘naturally’ have higher rates of entry.

An important aspect of entry is its impact on productivity growth through

innovation (Bartelsman et al., 2004). Although product innovation is less likely to

be a major driver of growth in a developing economy, there is still much innovation,

usually taking the form of an imitation of processes and technologies in developed

countries (Hausmann and Rodrik, 2003). As with process innovation in developed

economies, imitation in a developing economy is risky, for entrepreneurs do not

know a priori whether the new technology will be profitable in the economic and

2

institutional environment of the particular country. To capture these features we

follow Hausmann and Rodrick in modelling an industry from its inception in the

developing economy concerned.

Our model examines the effects of licence fees and bureaucratic delays in an

industry that is new to the given developing economy. We consider the initial level

of entry and the steady-state solution obtaining after any subsequent entry or exit.

In Section 2 we specify the characteristics of the industry, before considering in

Section 3 the free-market solution with and without licence fees, the latter being

denoted ‘laissez faire.’ Consistent with the empirical work cited above, we confirm

that higher licence fees can restrict entry, typically reducing both first-mover entry

and the steady-state number of firms in the industry. We also find that weaker

property rights can lead eventually to more entry and a higher level of production

because first movers into the industry are less able to extract rents from successful

innovation.

In Section 4 we introduce the idea that there can be delay in the granting of a

licence: a time lag between the payment of a licence fee and the award of the right

to enter. This has a significant impact on the behaviour of firms in the model. In

equilibrium, entrepreneurs choose to divide into three groups. One group pays the

fee immediately, entering after the required lag, while a second group never buys

the licence. Those in the third group choose to ‘speculate’ on a licence, buying the

3

licence early, before the profitability of early entrants is revealed. Depending on

the level of profitability that is revealed, some or all of the speculators may enter

without further delay, while any remaining speculators never enter. Interestingly,

the option of speculation causes all entrepreneurs to eschew the option of delaying

the decision as to whether to purchase a licence until profitability is revealed. If the

licence fee is large enough, however, entry by first movers is discouraged. Section

5 concludes.2

2 The Industry

Our modelling framework is deliberately parsimonious to establish that quite com-

plex results, in terms of interactions between first and second movers, and as a

consequence of institutional weaknesses, are inherent to even the simplest of entry

processes in a developing economy. The underdeveloped character of the economy

being analyzed is represented by factor supply constraints and institutional defi-

ciencies. To consider innovation in this context we follow Hausmann and Rodrick

(2003) by assuming that while innovation in developing countries will typically

be through the imitation of existing production methods in developed economies,

such technology is not common knowledge. Rather, the transfer of technology to

new economic and institutional environments requires adaptations, and there is an

associated uncertainty about the future of profitability of the new ventures. Hence,

4

even in a developing economy, when entrepreneurs set up firms in a particular ‘new’

industry, the profitability is initially unknown.

Consider a new modern-sector industry, with no incumbent firms at time t = 0.

Any entrepreneur may innovate, setting up a firm to enter the industry and produce

at t = 1. Though information on the supply of entrepreneurs is limited, studies of

self-employment and latent entrepreneurship (see, e.g., Blanchflower et al., 2001)

do not suggest that entrepreneurship is a function of income per capita (see also

Casson et al., 2006). Thus, the supply of potential entrepreneurs is assumed large

relative to the number that actually set up firms in the industry in equilibrium.3

Entrepreneurs (and firms) are indexed i = 1, 2, ...

Entry by entrepreneur i at t = 1 requires a sunk cost k in learning and setting

up in the industry and the payment of fee f ≥ 0 for a licence. If k and f are

incurred at t = 1, the firm will also need to employ a unit of skilled labour in any

period t in order to produce. The output of any active firm i at time t is

yit = θ, t = 1, 2, ... (1)

θ is the realization, at t = 1, of a stochastic variable Θ which is uniformly distrib-

uted with support [0, 2θ̄]. Given that at least one entrepreneur sinks the learning

cost k at t = 1, the realization of θ becomes common knowledge at t = 2. Θ

captures the idea that, although the industry may exist in other countries, its

5

suitability to local conditions and institutions can only be discovered by experi-

mentation; it represents uncertainty related to the quality and reliability of inputs

and their productivity under local climatic conditions. Note that θ is not firm-

specific. Unlike in Jovanovic (1982) or Ericson and Pakes (1995), entrepreneurs

do not learn about their own abilities; rather, they learn about their environment.

Apart from θ at t = 1, the values of all variables and parameters in the model are

common knowledge.

In this simplified framework, we assume that (for a high enough θ) the limit on

the profitability of production activity comes from an increasing supply price of

skilled labour, rather than the product demand side. We assume output demand

to be perfectly elastic, with price fixed at unity, so yit can also be interpreted as

revenue. In effect, we are assuming that the industry produces a traded good in a

small open economy. The wage wt per unit of skilled labour at time t is

wt = δ + αnt, δ, α > 0, t = 1, 2, ... (2)

where nt is the total number of firms in the industry at t.

Any number of entrepreneurs can enter the industry at any time. For a first

mover (that is, an entrant at t = 1) θ is stochastic. Then, for a potential second

mover (that is, an entrant at t = 2) the realization θ is known. We assume,

however, that although the production function (1), including θ, can be observed

6

by a second mover, methods of production can be only partially observed. A

second mover must therefore sink the learning/set-up cost (1 − γ)k, where 0 ≥

γ ≥ 1 is the spillover from the knowledge that a first mover acquires at t = 1.

These observations concerning second movers also apply to potential entrants at

t = 3, 4, ...

Since no information becomes available after θ is revealed at the beginning of

t = 2, there will be no reason for a firm to prefer to enter or exit in later periods,

rather than at the beginning of t = 2. Therefore, for t ≥ 3, nt = n2, yit = yi2, and

wt = w2. Writing πit and πjt for the respective profits at time t of a first mover i

and a second mover j, we have

πi1 = θ − δ − αn1 − k − f ; πit = θ − δ − αn2, t = 2, 3, ...;

πj2 = θ − δ − αn2 − (1− γ)k − f ; πjt = θ − δ − αn2, t = 3, 4, ... (3)

The present value, at t = 2, of first mover i’s profit stream, if it stays in production,

is

V i2 =

1

1− σ(θ − δ − αn2) , (4)

where σ ∈ (0, 1) is a discount factor. The present value, at t = 2, of second mover

j’s profit stream is

V j2 =

1

1− σ(θ − δ − αn2)−K, (5)

7

where

K ≡ (1− γ)k + f . (6)

In each case that follows, we examine the entry by first movers, the subsequent

pattern of entry and exit, and the steady-state number of firms n∗ in the industry.

n∗ is the number of firms in the industry after uncertainty has been resolved and

any adjustments associated with licence delay have played out (n∗ is the number

of firms that enter, but do not exit).

3 Market Equilibrium

Assume that entrepreneurs make decisions freely about whether to enter or exit.

In this section we assume that although a licence may be required, there is no delay

in receiving it. Since there is no uncertainty after θ is revealed (between t = 1 and

t = 2), the equilibrium number of firms is the same at t = 3, 4, ... as at t = 2. We

therefore focus on the values of n1 and n2, and we solve the model by backward

induction from t = 2. At t = 2, each entrepreneur maximizes the present value

of his or her profit stream, given θ. Taking into account the behaviour that will

obtain for each entrepreneur at t = 2 for all possible θ, at t = 1 each entrepreneur

maximizes the expected present value of his or her profit stream. We assume that

parameter values are such that there is positive entry at t = 1 in the solution.4

The value of n2 depends on the range of values within which θ falls. Four cases

8

can be distinguished.

Case (a) All n1 first movers would exit because even one of them alone in the

industry would make a loss at t = 2. From (3), this occurs if

θ < δ + α ≡ θa. (7)

If (7) holds, all first movers exit; and since second movers are at a cost disadvantage

relative to first movers (still having to incur a set-up cost), there are is no entry

by second movers.

Case (b) In this range some of the n1 first movers remain in the industry at

t = 2 (thus, θ ≥ θa) but some exit. If all n1 first movers were to remain in the

industry the n1-th would make a loss, i.e., from (3),

θ < δ + αn1 ≡ θb(n1). (8)

Thus, in this case θ ∈ [θa, θb). Because, at t = 2, a potential second mover is at a

cost disadvantage relative to any first mover, there is no entry by second movers.

The number of firms n2 in this case is such that V i2 = 0, so that, from (4),

n2 =θ − δ

α> 0 for θ ∈ (θa, θb]. (9)

Case (c) All first movers remain in the industry at t = 2 (thus, θ ≥ θb) but θ

9

is not so great as to induce entry by second movers. From (5), V j2 is decreasing in

n2. Thus, if Vj2 ≤ 0 for n2 = n1, there will be no second movers. This condition

can be written

θ ≤ δ + αn1 + (1− σ)K ≡ θc(n1). (10)

When θ ∈ [θb, θc), n2 is independent of parameter values.

Case (d) If θ ∈ (θc, 2θ̄] the first movers will all remain in the industry at t = 2

and there will be entry by second movers until V j2 = 0. Hence, from (5),

n2 =1

α[θ − δ − (1− σ)K] ≡ n̂2. (11)

At t = 1 firm n1 is the marginal firm among those that enter. Writing πn1t for

its profit at time t, and taking into account the four cases for possible draws of θ,

the expected present value of this firm’s profit stream is

V n11 (n1) =

1

2θ̄

(Z 2θ̄

0

πn11 (n1)dθ +σ

1− σ

"Z θc

θb

πn12 (n1)dθ +

Z 2θ̄

θc

πn12 (n̂2)dθ

#).

(12)

Here, the term in parentheses (.) after each πn1t denotes the number of firms in the

industry in the time period considered. The first integral covers profit at t = 1,

while the second and third integrals relate to profit at t = 2 in Cases (c) and

(d) respectively. (If Case (a) or (b) applied, firm n1 would exit, and so no term

is specified.) For (12) to hold we assume that 2θ̄ > θc; i.e., we assume that the

10

realization of θ may be sufficiently large for there to be second-mover entry. For

simplicity, we do not consider other cases. πn11 (.) and πn12 (.) are given by (4), while

n̂2 is given by (11). Thus,

V n11 (n1) = (θ̄ − δ − αn1 − k − f)− 1

4θ̄σ(1− σ)K2 + σK

µ1− δ + αn1

2θ̄

¶. (13)

From (13), dV n11 (n1)/dn1 < 0.

5 n1 adjusts such that, in equilibrium, Vn11 (n1) =

0, the solution being n1 = n̂1(f), where

n̂1(f) =2θ̄ − δ

α− 4θ̄(θ̄ + k + f) + σ(1− σ)K2

2α(2θ̄ + σK)≡ 2θ̄ − δ

α− z. (14)

We assume that the configuration of parameter values is such that n̂1(f) > 0.

Combined with our assumption that 2θ̄ > θc,6 using (10) and (14), this implies

that

2θ̄ − δ > αz > (1− σ)K. (15)

From (11) and (14), we can find, for Case (d), how many second movers will enter.

Denoting this number by m̂2(f) = n̂2(f)− n̂1(f), we obtain

m̂2(f) =4θ̄(θ − θ̄ + γk) + σK [2θ − (1− σ)K]

2α(2θ̄ + σK)for θ ∈ (θc, 2θ̄]. (16)

Let n∗m(f) denote the steady-state number of firms in the market solution for

11

licence fee f . Since by t = 2 all uncertainty is resolved, n∗m(f) = n2. The value

of n∗m(f) depends on which of cases (a)-(d) obtains.

Our first proposition gives the comparative statics of the model. These are in-

tuitively as might be expected, with the initial number of entrants and the steady-

state number of firms non-decreasing in the discount factor and non-increasing in

cost parameters. We consider the role of f separately below, but one other result

is worth commenting on here. A higher value of the spillover parameter γ is associ-

ated with a smaller entry of first movers, but, provided θ is large enough for second

movers to enter, this effect is more than outweighed (after a one-period lag) by the

greater entry of second movers. This has the implication that, if weaker property

rights are associated with greater levels of spillover, they will cause a lower level

of entry initially, but competitive pressures can lead eventually to more entry in

total, with a larger number of new firms and a higher level of output. Pressures

in developing economies to strengthen intellectual property rights, for example via

the World Trade Organization, in so far as they affect the level of spillovers in

the domestic market, may have the perverse effect of reducing the total number of

firms and production.

Proposition 1 n̂1 is increasing in σ, and decreasing in α, δ, γ, k and f . In Case

(a) n∗m = 0. In Case (b) n∗m is decreasing in α and δ, and independent of σ, γ,

k and f . In Case (c) n∗m = n̂1. In Case (d), n∗m is increasing in σ and γ, and

12

decreasing in α, δ, k and f .

Setting f = 0, we obtain the laissez-faire solution, which is depicted in Figure

1, where period-1 entry is labelled nm1 (0) and the steady-state solution is labelled

n∗m(0). For at least one firm to stay in the industry we must have θ ≥ δ + α.

nm1 (0) is independent of θ because it is determined before θ revealed. The vertical

distance between n∗m(0) and nm1 (0) shows the amount of entry at t = 2. For low

values of θ this is negative, i.e., there is exit.

[Figure 1]

Figure 2 illustrates the effect of a positive licence fee on the market solution.

The zero-licence fee case is shown by the broken line, and the positive-licence fee

case is shown by the thick line. As indicated in Proposition 1, the existence of

the licence fee reduces the number of first movers, and so the horizontal section of

n∗m(f) is below that of n∗m(0). Also, provided θ > θb(n̂1), the licence fee reduces

the steady-state number of firms.7

[Figure 2]

4 Delay and Speculation

In the formulation above, the licence fee f plays almost the same role as the

learning cost k, the only difference being that while all firms pay the same fee, the

13

existence of a spillover causes second movers, though not first movers, to pay only

a portion of k. We now suppose, however, that there is a one-period delay between

paying the licence fee and getting the licence. Thus, we suppose that first movers

pay the licence fee at t = 0 and begin production at t = 1. As above, θ is observed

between t = 1 and t = 2. With this amendment to the model, one possibility is

for other firms to wait to observe θ and then, if this realization is favourable, pay

the fee f immediately. Thus, f would be paid at the beginning of t = 2 and then

entry by second movers would occur at the beginning of t = 3, with the learning

cost (1− γ)k then being incurred.

However, entry may take a different form. An entrepreneur may decide to

‘speculate,’ applying for a licence, but not producing before θ is realized, and then

only going into production if the realization is sufficiently favourable. Thus, the

entrepreneur may pay the fee at the beginning of t = 1, and then either begin

production at the beginning of t = 2 or not begin production at all. We can

therefore now distinguish three types of entry: by first movers, by speculators and

by ‘later movers’ (entrepreneurs who wait to see the realization θ before possibly

paying f).

At t = 2 a first mover has a cost advantage (1−γ)k over a speculator, and so, if

a speculator enters, we know that all first movers are remaining in the industry. At

t = 2 and t = 3 a speculator has the cost advantages of f and (1−γ)k over a later

14

mover, and so, if any later movers enter, we know that all speculators are staying

in. To solve the model, we begin by disregarding later movers entirely, considering

only first movers and speculators. We shall then show that no entrepreneurs will

choose to be later movers.

Suppose that s1 entrepreneurs buy licences at t = 1. For any one of these

speculators, h, if he or she then enters at t = 2, profit is

πh2 = θ − δ − αn2 − (1− γ)k;

πht = θ − δ − αn3, t = 3, 4, ... (17)

And, at t = 2, the present value of its profit stream is

V h2 = θ − δ − αn2 − (1− γ)k +

σ

1− σ(θ − δ − αn3) . (18)

The number of speculators that then enter at t = 2 depends on what realization

of θ occurs. The realizations can be divided into three cases.

Case (SA) If θ is low enough, V h2 < 0 for all h ∈ (n1, n1 + s1], so that none of

the speculators enter. Since, at t = 2, a speculator has a cost disadvantage relative

to a first-mover, the highest value of θ at which this case obtains is when all n1

first movers would nonetheless remain in the industry. Hence, Case (SA) is defined

15

by writing n2 = n3 = n1 in (18) and finding the values of θ for which V h2 < 0, i.e.,

θ < δ + αn1 + (1− σ)(1− γ)k ≡ θSA(n1). (19)

Case (SB) In this range (19) is violated, and some, but not all, s1 speculators

enter. If all speculators were to enter, the least efficient would make a loss in

present value terms; i.e., from (18),

θ < δ + α(n1 + s1) + (1− σ)(1− γ)k ≡ θSB(n1 + s1). (20)

The number of firms adjusts such that in (18), V h2 = 0 for h = n2, i.e.,

n2 =1

α[θ − δ − (1− σ)(1− γ)k] ≡ n̂s2. (21)

Case (SC) Here, all s1 speculators enter; i.e., (20) is violated.

Moving back to t = 1, we can now consider the payoff from speculation. The

expected present value for the marginal speculator (the s1-th) is

V s11 (n1 + s1) = −f + σ

2θ̄

Z 2θ̄

θSB

·πn1+s12 (n1 + s1) +

σ

1− σπn1+s13 (n1 + s1)

¸dθ.

The first term in [.] is profit including set-up cost (1− γ)k, while the second is the

16

stream of discounted profits after the set-up cost has been incurred. Hence,

V s11 (n1 + s1) = −f + σ

θ̄(1− σ)

µθ̄ − θSB

2

¶2. (22)

From (20) and (22), dV s11 (n1 + s1)/ds1 < 0. s1 adjusts such that, in equilibrium,

V s11 (n1 + s1) = 0, that is, (20) and (22),

n1 + s1 =1

α

(2θ̄ − δ − (1− σ)(1− γ)k − 2

·θ̄(1− σ)f

σ

¸ 12

). (23)

The ranges of θ relevant to the behavior of a first mover follow immediately

from the cases already specified in this and the previous section.

Case (Fa) All first movers exit (and no speculators enter). They do this if

θ < θa, as specified in (7).

Case (Fb) Some, but not all, first movers exit (and no speculators enter). This

occurs if θ ≥ θa, but θ < θb(n1), as specified in (8). Note that since the value of n1

will now be different to the value taken in the absence of delay, the value of θb(n1)

will also differ. n2 is now given by (9).

Case (Fc) All first movers stay in production, but still no speculators enter.

In this case θ ≥ θb(n1), but θ < θSA(n1), where θSA(n1) is given by (19).

Case (Fd) All first movers stay in production and some, but not all, speculators

enter. This occurs if θ ≥ θSA(n1), but θ < θSB(n1), where θSB(n1) is given by

17

(20). n2 is now given by (21).

Case (Fe) All speculators enter. This happens if θ ≥ θSB(n1), the number of

firms n1 + s1 being given by (23).

Given these ranges, the present value, measured from t = 0, of the expected

profit stream for the marginal first mover is

V n10 (n1) = −f + σ

2θ̄

Z 2θ̄

0

πn11 (n1)dθ +

σ2

1− σ

1

2θ̄

"Z θSA

θb

πn12 (n1)dθ +

Z θSB

θSA

πn12 (n̂s2)dθ +

Z 2θ̄

θSB

πn12 (n1 + s1)dθ

#,

where the profit equations (3) apply, but with f deleted. Note that, using (20) and

(23), it is found that 2θ̄ > θSB, and so the final integral is valid. The first integral

covers profit at t = 1; the others cover profit at t = 2, 3, ... for Cases (Fc-Fe), i.e.,

when θ is large enough for the marginal first mover to stay in production.

We thus find, after substituting from (23) to eliminate s1, that

V n10 (n1) = −(1− σ)f + σ(θ̄ − δ − αn1 − k)− 1

4θ̄(1− σ) [σ(1− γ)k]2

+σ2(1− γ)k

µ1− δ + αn1

2θ̄

¶. (24)

From (24), dV n10 (n1)/dn1 < 0. n1 adjusts so that V

n10 (n1) = 0, the solution being

18

n1 = n̂s1, where, from (24),

n̂s1 =2θ̄ − δ

α− 4θ̄

£(θ̄ + k) + (1− σ)f/σ

¤+ (1− σ)σ [K − f ]2

2α£2θ̄ + σ(K − f)

¤ . (25)

We now add the possibility of later entry into the model, while still allowing

for speculation. If entrepreneur j pays at t = 2 for a licence in order to begin

production at t = 3, the present value of his or her profit stream is

V j2 =

σ

1− σ(θ − δ − αn2)− σ(1− γ)k − f . (26)

This parallels equation (5), but incorporates additional discounting because of the

time lag in receiving the licence. If θ were to be sufficiently high for such late

entry to occur, an interior solution for the number of firms n3 at t = 3 would be

characterized by V n32 = 0; and hence,

n3 =1

α

½θ − δ − (1− σ)

·(1− γ)k +

f

σ

¸¾. (27)

However, n3 is the total number of firms in this case; that is, it includes the

first movers and speculators, as well as later movers. Now compare (27) with (21),

which gives the total number of firms in Case (SB), i.e., when the marginal spec-

ulator does not enter, though some speculators do. Because of the appearance of

the term f/σ, n3 in (27) is less than n2 in (21). But when there is any late entry,

19

our assumptions imply that all first movers have stayed in and all speculators have

entered. Thus, the total number of first movers and speculators must be at least

as large as in (21). We have a contradiction: (27) could only apply if the num-

ber of later movers were negative. The following proposition therefore obtains.8

Proposition 2 With a positive licence fee, if it is necessary to wait to be granted a

licence, the speculative purchase of licences is undertaken by entrepreneurs to such

an extent that the later (non-speculative) purchase of licences is entirely eliminated.

The up-front payment of the licence fee by speculators gives them a strategic

advantage over later entrants. We can therefore disregard the possibility of later

entry when there is a delay in being granted a licence. The pattern of entry is as

depicted in Figure 3, where ns1(f) is period-1 entry in the delay-case. The pattern

in Figure 3 has the same general shape as that in Figure 1, except in one important

respect. For θ ∈ [θSB(n1), 2θ̄], the line depicting the steady-state number of firms

n∗ is horizontal. This is because, for such high θ all the entrepreneurs that have

speculatively bought a licence will enter, and, as we have seen, the solution to the

model is such that there is no entry by later movers.

[Figure 3]

Finally, for f > 0, we compare the solutions with and without delay. Given

20

Proposition 1, we compare (25) with (14) to obtain the following result for first

movers.

Proposition 3 There exists a critical value f = f̄(σ) such that, if (1 − γ)(2 −

3σ)k + (2 − σ)[(1 − σ)f + δ] > 0, then n̂s1 R n̂1 as f Q f̄(σ); i.e., for f < f̄(σ)

licence delay increases first-mover entry, and vice versa.

Proof. From (14) and (25), using (6), we obtain 2α(2θ̄ + σK)[2θ̄ + σ(K −

f)](n̂s1− n̂1)/f = 2θ̄[2θ̄(4σ− 2−σ2)/σ− 2σγk+σ(1−σ)(1 −γ)k] −(1−σ)K[(2−

σ)2θ̄ − σ2(1 − γ)k] ≡ A. Substituting from (6), it is found that A R 0 as f Q

2θ̄[2θ̄(4σ−2−σ2)/σ−2σγk−(1−σ)2(1−γ)k]+(1−σ)(1−γ)kσ2(1−γ)k(1−σ)[(2−σ)2θ̄−σ2(1−γ)k] , provided 2θ̄ > σ2(1− γ)k/(2−

σ). From (15), 2θ̄ > (1 − σ)K + δ, and so a sufficient condition for 2θ̄ >

σ2(1 − γ)k/(2 − σ) is that (1 − σ)K + δ ≥ σ2(1 − γ)k/(2 − σ). Using (6), this

reduces to (1− γ)(2− 3σ)k + (2− σ)[(1− σ)f + δ] > 0.

Delay benefits first movers here in that potential competitors (speculators) pay

the fee for a licence that they may not use, and this limits their expected profit.

But the cost to the potential first mover is that it must itself wait, after payment

for the licence, before entry is possible. Because of the latter effect, we find that

delay induces greater first-mover entry only if f is sufficiently small. Indeed, we

cannot rule out the possibility that f̄(σ) < 0, in which case licence delay necessarily

causes smaller first-mover entry.

Denoting the steady-state number of firms in the presence of licence delay by

21

n∗m(f), we can compare this with the corresponding number n∗m(f) without delay.

The relative size of these numbers depends on the realization θ and on parameter

values, but we can make some comments by reference to Figure 3, where both are

shown. Starting from θ = 0, and then increasing θ, the steady state is at first zero

in each case, and then the same upward-sloping lines apply; but then, since we

cannot determine, unconditionally, in which case there is higher first-period entry,

we cannot tell which steady-state number is higher (we cannot tell in which case

the horizontal portion of the steady-state line in Figure 3 is higher; the figure is

merely an illustration). However, for the highest θ, provided f < 4θ̄/σ(1−σ), i.e.,

provided f is not so high as to prevent entry in the no-delay case, n∗m(f) > n∗s(f),

as drawn.

5 Conclusions

Entry rates in developing economies are potentially high because numerous oppor-

tunities exist to close the technology gap with developed economies, including by

the imitation of processes and technologies. We explore the process of entry in a

model in which two of the entry barriers that are widely observed in developing

economies - licence fees and bureaucratic delay - may in principle prevent these

opportunities from being exploited. We find that the effects vary significantly

across barriers. A licence fee alone reduces both the number of first movers and

22

the steady-state number of firms, but in combination with bureaucratic delay can

have the opposite effects. This is because, although the delay between application

and receipt of a licence is assumed the same for all firms, delay changes the balance

of advantage between first and later movers.

Our model opens up an interesting new line of empirical inquiry. It suggests

that, in an institutional setting in which bureaucratic delays are a barrier to entry,

we may expect that in a new industry entrepreneurs will purchase more licences

being than they then use. Data are not currently available to investigate the extent

of such speculative behaviour in practice, but we propose a data-gathering exercise

to test this hypothesis empirically.

Notes

1Licence fees are common institutional features in both developed and devel-

oping economies, but bureaucratic delays are typically significantly longer in de-

veloping economies (Djankov et al, 2002).

2We do not use our model to provide a normative analysis because fees and

delays are generally not manipulated consciously by bureaucrats as industrial pol-

icy tools to achieve welfare ends in practice, and in any case are unlikely to be

the most appropriate policies for this purpose. However, in an earlier version of

the paper, Bennett and Estrin (2006), we characterize what the pattern of entry

23

would be if it were controlled by a welfare-maximizing social planner. We refer to

this analysis in note 7 below.

3At any time t there are more potential entrants than the market can sustain.

The question therefore arises of how the entry of only some of these potential

entrants is co-ordinated. We assume that there is some small exogenous asymmetry

between potential entrants that allows entry to take place only up to the point at

which the present value of the expected profit stream for the marginal entrant is

zero. An alternative approach would be to model a mixed-strategy equilibrium,

but this would add complexity, and Levin and Peck (2003) have shown that this

approach can generate some rather implausible results.

4If all firms find it unprofitable to enter at t = 1, the realization of θ will not be

revealed, and so they will also find it unprofitable at t = 2, 3, ... and the industry

will not be established.

5Throughout, we approximate by treating the number of firms as continuous.

6The two assumptions are required for the number of first movers to be positive

(otherwise there is no industry) and for the possibility that profitability can turn

out sufficiently high for second movers to enter.

7In Bennett and Estrin (2006) we show that laissez faire leads to excessive

entry at t = 1 and, provided θ > δ + α, there is is an excessive number of firms in

the laissez-faire steady state. This is because each entrepreneur fails to take into

24

account that he or she is sinking costs for learning to an extent that is unnecessary

from a social point of view. As a result, the introduction of a relatively small

licence fee raises welfare, but the welfare implications of delay are less clear-cut.

8The arguments in this section are based on licence delays when a positive fee

exists. In the absence of a fee f all firms would procure a licence at t = 0 and we

would effectively be back in the laissez-faire solution.

References

[1] BARTELSMAN, E. J., HALTIWANGER, J. and SCARPETTA, S. (2004),

Microeconomic evidence of creative destruction in industrial and developing

countries, Tinbergen Institute Discussion Papers, 04/078/3.

[2] BENNETT, J. and ESTRIN, S. (2006), Regulatory barriers and entry in

developing economies, CEDI Working paper 06-02, Brunel University.

[3] BLANCHFLOWER, D. G., OSWALD, A. and STUTZER, A. (2001), Latent

entrepreneurship across nations, European Economic Review, 45, 680-691.

[4] CASSON,M., YEUNG, B., BASU, A. andWADESON, N. (2006), The Oxford

Handbook of Entrepreneurship, Oxford: Oxford University Press.

25

[5] CAVES, R. E. (1998), Industrial organization and new findings on the

turnover and mobility of firms, Journal of Economic Literature, 36, 1947-

1982.

[6] DE PAULA, A. and SCHEINKMAN, J. (2006), The informal sector, UCLA

Department of Economics working paper.

[7] DESAI, M., GOMPERS, P. and LERNER, J. (2003), Institutions, capital

constraints and entrepreneurial firm dynamics: evidence from Europe, NBER

Working paper No. w10165.

[8] DE SOTO, H. (1990), The Other Path, New York, NY: Harper and Row.

[9] DJANKOV, S., LA PORTA, R., LOPEZ-DE-SILANES, F. and SHLEIFER,

A. (2002), The regulation of entry, Quarterly Journal of Economics, 117, 1-35.

[10] DUNN, T., ROBERTS, M. and SAMUELSON, L. (1989), The growth and

failure of US manufacturing plants, Quarterly Journal of Economics, 104,

671-698.

[11] ERICSON, R. and PAKES, A. (1995), Markov-perfect industry dynamics: a

framework for empirical work, Review of Economic Studies, 62, 1, 53-82.

[12] EVANS, D.S. (1987), Tests of alternative theories of firm growth, Journal of

Political Economy, 95, 657-674.

26

[13] GEROSKI, P. (1995), What do we know about entry? International Journal

of Industrial Organization, 13, 421-440.

[14] HAUSMANN, R. and RODRIK, D. (2003), Economic development as self

discovery, Journal of Development Economics, 72, 603-33.

[15] JOVANOVIC, B. (1982), Selection and the evolution of industry, Economet-

rica, 50, 649-70.

[16] KLAPPER, L., LAEVEN, L. and RAJAN, R. G. (2004), Business environ-

ment and firm entry: evidence from international data, CEPR Discussion

Paper No. 4366, April.

[17] LEVIN, D. and PECK, J. (2003), To grab the market or bide one’s time: a

dynamic model of entry, RAND Journal of Economics, 34, 536-556.

[18] MALONEY, W. F. (2004), Informality revisited, World Development, 32,

1159-78.

[19] ROBERTS, M. J., and TYBOUT, J. R., ed. (1996), Industrial Evolution in

Developing Countries: Micro Patterns of Turnover, Productivity, and Market

Structure, New York: Oxford University Press.

[20] SCARPETTA, S., HEMMINGS, P., TRESSEL, T., and WOO, J. (2002), The

role of policy and institutions for productivity and firm dynamics: evidence

27

from micro and industry data, Economics Department, OECD, Working Pa-

per No. 329.

[21] TYBOUT, J. R. (2000), Manufacturing firms in developing countries: how

well do they do, and why? Journal of Economic Literature, 38, 11-44.

28

Figure 1. Laissez Faire (f = 0) nt

n*m(0) n1

m(0)

Ο 2θ θ

Figure 2. A Positive Licence Fee

nt n*m(0) n*m(f) Ο 2θ θ

Figure 3. Speculation nt

n*s(f) n1

s(f)

no delay: n*m(f)

Ο 2θ θ