- Email: [email protected]
A credit market in early stages of economic development
Economics Letters 112 (2011) 42–44
Contents lists available at ScienceDirect
Economics Letters journal homepage: www.elsevier.com/locate/ecolet
A credit market in early stages of economic development Hideki Nakamura ∗ , Tetsuya Nakajima Faculty of Economics, Osaka City University, 3-3-138 Sugimoto, Sumiyoshi, Osaka 558-8585, Japan
article
info
Article history: Received 16 March 2010 Received in revised form 9 February 2011 Accepted 18 March 2011 Available online 6 April 2011
abstract Even if the relatively rich and the poor are initially caught in a poverty trap, the relatively rich can escape poverty by receiving payments from the poor. Further accumulation of wealth by the rich allows the poor to escape poverty. © 2011 Elsevier B.V. All rights reserved.
JEL classification: D91 O15 O16 Keywords: Credit market Early stages of economic development Relatively rich Poor
1. Introduction Models explaining the persistence of poverty and inequality are typically based on a poverty-trap mechanism—individuals with wealth above some threshold converge to a high-income steady state and those below the threshold converge to a low-income steady state.1 However, one or two centuries ago, most of the ancestors of the relatively wealthy of today had little wealth. How did they escape poverty? By considering a trickle-down effect, Aghion and Bolton (1997) and Galor and Moav (2004) examined the dynamics of wealth distribution and showed that the poor could escape poverty when the rich continue accumulating their wealth. If not only the poor, but also the relatively rich are caught in a poverty trap, the economy cannot take off. Considering a credit market in early stages of economic development, this paper presents a theory that allows the relatively rich to accumulate wealth. Following Moav (2002) and Galor and Moav (2004), a poverty trap is driven by increasing saving rates in the form of transfers to the offspring. Poor families then converge to a corner solution of no transfer. However, the relatively rich – individuals with some wealth – accumulate more wealth by lending to the
∗
Corresponding author. Tel.: +81 6 6605 2271; fax: +81 6 6605 3066. E-mail address: [email protected] (H. Nakamura).
1 See, for example, Galor and Zeira (1993) which is pioneering work. 0165-1765/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.econlet.2011.03.011
poor, who pay a high interest rate as a result of limited supply of funds and a high demand. The relatively rich – once credit markets are well functioning – could accumulate wealth and escape a lowincome trap before the poor. Galor (2005) extensively surveys long-run transition models offering explanations for the takeoff from poverty to growth. This paper presents a theory that allows the relatively wealthier, even if poor according to modern standards, to accumulate wealth, and, thereby, change the properties of the dynamic system over time. Furthermore, as found in Demirgüc-Kunt and Levine (2009), financial improvement has a significant positive effect on the income of the poor because it boosts efficiency and the equality of opportunity. This paper highlights the importance of improvements in the credit market for the process of escaping poverty.2 First, it allows the relatively rich to become richer. Second, further improvements in credit markets and the accumulation of wealth by the rich allow for a reduction in the interest rate and a trickle-down effect. The poor then can escape poverty.3
2 In the US, many paid workers appeared in urban areas after the Civil War. Their wages were low and their employment was unstable, and they had almost no wealth. This prompted the emergence of credit markets. Furthermore, Morris Plan banks, the first one being established in 1910, contributed to increased credit availability by lowering interest rates. See Mushinski and Phillips (2008). 3 In some less developed countries, credit markets are not functioning well. For example, in Nepal, many households that have worked as bonded laborers in the
H. Nakamura, T. Nakajima / Economics Letters 112 (2011) 42–44
43
Taking (1) into account, the dynamics of investment in the case of (5) become eit = (1 − α)f (eit −1 ) − αθ .
(7)
Curve C in Fig. 1 depicts this equation. We consider the case that (7) has two stationary states: e∗ and e∗∗ . In Fig. 1, we assume that
(1 − α)f ′ e∗ < 1 and (1 − α)f ′ e∗∗ > 1. (8) Let us consider the case in which bp,−1 = 0 and 0 < br ,−1 <
e∗∗ . The bequest level of the poor remains zero. Because the initial wealth level of the relatively rich is below the threshold, it decreases and becomes zero after some time. Thus, economic development never occurs. 3. Model with a credit market Now, suppose that a credit market becomes available. Eqs. (5) and (6) are then modified as
Fig. 1. Dynamics of the investment level.
bit = (1 − α)Iit − αθ ,
2. Basic model with no credit market
bit = 0,
if (1 − α)Iit > αθ ,
(9)
if (1 − α)Iit ≤ αθ ,
(10)
Our model is a closed overlapping-generations economy. If parents decide to leave their bequests, their children receive these bequests in the first period. In the second period, they produce a single type of goods. We first consider the case in which there is no credit market. The output can be used for either consumption or bequest. A bequest is used for investment. Individuals are identical except for the bequests that are received from their parents. We assume that the initial numbers of rich and poor are ηL and (1 − η)L, respectively, where L represents the population of each generation. Each individual born in period t − 1 receives a bequest from their parent and spends it just for investment, because there is no credit market. That is, we have
where Irt and Ipt are the respective income levels of the relatively rich and the poor in period t. An individual determines the optimal investment level to maximize their income by borrowing or lending. The incomemaximization problem of the relatively rich, who are the lenders, is
eit −1 = bit −1 ,
brt = (1 − α) f (ert −1 ) + f ′ (ert −1 ) (brt −1 − ert −1 ) − αθ .
(1)
where i = r , p; brt −1 and bpt −1 are the respective bequest levels of the relatively rich and the poor that are received in period t − 1; and ert −1 and ept −1 are the respective investments in period t − 1. In period t, each individual produces goods subject to the following production function yit = f (eit −1 ),
(2)
where yrt and ypt are the respective levels of outputs in period t. We assume that f ′ (eit −1 ) > 0 and f ′′ (eit −1 ) < 0. We also assume that f (0) > 0. We assume the altruistic bequest motive, i.e., the ‘joy of giving’. A zero bequest is allowed because individuals can live with no investment. The utility maximization of an individual born in period t − 1 is as follows max α ln cit + (1 − α) ln(bit + θ ), cit ,bit
0 < α < 1, θ > 0,
(3)
max Irt = f (ert −1 ) + rt (brt −1 − ert −1 ) ,
(11)
ert −1
where rt is the gross interest rate in the credit market. The investment level is determined as f ′ (ert −1 ) = rt .
(12)
Eq. (9) is rewritten as
(13)
On the other hand, the poor, who have no wealth, become the borrowers. Because the lenders to individuals have positive costs of keeping track of each borrower, the individual must borrow at a rate higher than rt . They solve max Ipt = f ept −1 − δ rt ept −1 ,
ept −1
δ ≥ 1.
(14)
The investment level of the poor is lower than that of the relatively rich because of an imperfect credit market f ′ ept −1 = δ rt .
(15)
Given the interest rate, the cost of keeping track increases with an increase in the investment level. Borrowers must pay higher interests as δ rises. 4. Can the credit market help the relatively rich escape poverty? The credit market equilibrium is given by
s.t. cit + bit = f (eit −1 ),
(4)
η(brt − ert ) = (1 − η)ept .
(16)
where crt and cpt are the respective consumption levels. We assume that (1 − α)f (0) < αθ . The first-order conditions yield
The investments of the relatively rich and the poor are financed by the wealth of the relatively rich only. We specify the production function as
bit = (1 − α)f (eit −1 ) − αθ ,
(5)
f (eit ) = a0 + a1 eit ,
(6)
Given (17), the investment level of the poor is proportionate to that of the relatively rich
bit = 0,
if (1 − α)f (eit −1 ) > αθ ,
if (1 − α)f (eit −1 ) ≤ αθ .
When the income level is low, the bequest level becomes zero because of the shift parameter, θ .
γ
ept = gert ,
(17)
(18) −1/(1−γ )
tea industry manage their living expenditure by further borrowing at high interest rates.
a0 , a1 > 0, 0 < γ < 1.
where g ≡ δ ≤ 1. While an increase in g implies an improvement in credit availability, the difference in investment levels between the relatively rich and the poor becomes small. Using (13), (16) and (18),
44
H. Nakamura, T. Nakajima / Economics Letters 112 (2011) 42–44
the dynamics of the investment level can be represented as ert =
1−α
ηf (ert −1 ) + (1 − η) gf ′ (ert −1 ) ert −1
η + (1 − η) g ηαθ − . η + (1 − η) g
(19)
The relatively rich obtain not only income from their own investment, but also the interest payments from the poor that increase with an improvement in the degree of credit availability. Eq. (19) is drawn as curve D in Fig. 1. We assume that curve D is strictly concave. In Fig. 1, eD indicates a stable stationary state, and eDD is an unstable stationary state. Curve D approaches curve C as g decreases.4 We define eˆ as an investment that satisfies
(1 − α) f eˆ − f ′ eˆ eˆ − αθ = 0.
(20)
Curves C and D intersect at the point where ert −1 = eˆ regardless of the degree of credit market imperfection. How does an improvement in credit availability affect the investment levels of stationary states? The comparative statistics with respect to g are
where
{(1−α)[f (e)−f ′ (e)e]−αθ } = − 1−[(1−α)/(η+(1−η)g )]{[η+(1−η)g]f ′ (e)+(1−η)gf ′′ (e)e} . η(1−η)/(η+(1−η)g )2
An improvement in credit availability decreases the investment levels of the relatively rich on the stationary states. When credit availability is improved, the wealth of the relatively rich is more efficiently allocated between the relatively rich and the poor. The concavity of the production technology implies that the investment level of the relatively rich decreases while the level for the poor increases. The income and wealth levels of the relatively rich on the high-level stationary state increase because of an increase in the interest payments from the poor to the relatively rich. The income level of the poor on the stationary state also increases because of an increase in their investments. It is possible for the relatively rich to escape poverty because of the existence of a credit market. The relatively rich can accumulate their wealth if the initial investment level is greater than eDD , which is lower than the threshold e∗∗ , i.e., eDD < er ,−1 =
η br ,−1 < br ,−1 < e∗∗ . η + (1 − η) g
5. Conclusion We considered an economy in the early stages of economic development, where not only the poor but also the relatively rich are initially caught in poverty. We found that a credit market potentially has the power to help the relatively rich accumulate their wealth before the poor.
∂ eD ∂ eDD < 0 and < 0, ∂g ∂g ∂e ∂g
In the early stages of economic development, many individuals have no wealth to invest. A few individuals who have small wealth exist. However, investing all their wealth in their own production cannot yield enough income to escape the low-income trap, because production is subject to diminishing returns. If a credit market is available, the relatively rich lend a part of their wealth to the poor. The relatively rich can receive a large amount of interest payments from the poor as a result of limited supply of funds and a high demand. The relatively rich can then escape the trap by accumulating their wealth. Thus, a credit market with a sufficient degree of credit availability can make an economy take off because of the resource-allocation effect and the incomedistribution effect. In addition, we examine whether the poor can escape poverty. If the investment level of the poor exceeds eˆ , defined in (20), until the investment level of the relatively rich converges to eD , the poor can start accumulating their own wealth. There exists g˜ that satisfies g˜ eD = eˆ . Therefore, there exists a trickle-down effect if a credit market is sufficiently improved.
(21)
We also obtain ∂ br /∂ g < 0 at e = eDD , where br = eDD [η/(η + (1 − η)g )]−1 . Thus, it becomes easy for the relatively rich to escape poverty by improving credit availability.5 Here, we have the following proposition. Proposition 1. Given (21), a credit market can help the relatively rich, who are caught in a poverty trap, to escape the trap.
4 The proof, which shows that eDD < e∗∗ < eD < e∗ , is available on request. 5 Furthermore, (21) can hold when the ratio of the relatively rich to the total population is small.
Acknowledgments We would like to thank Yoshihiko Seoka, Mario Oshima, and the seminar participants of Osaka University of Economics for their comments. A referee provided helpful suggestions to improve the paper. References Aghion, P., Bolton, P., 1997. A theory of trickle-down growth and development. Review of Economic Studies 64, 151–172. Demirgüc-Kunt, A., Levine, R., 2009. Finance and inequality: theory and evidence. NBER Working Paper Series 15275. Galor, O., 2005. From stagnation to growth: unified growth theory. In: Aghion, P., Durlauf, S. (Eds.), Handbook of Economic Growth, vol. 1A. Amsterdam, NorthHolland, pp. 171–293. Galor, O., Moav, O., 2004. From physical to human capital accumulation: inequality and the process of development. Review of Economic Studies 71, 1001–1026. Galor, O., Zeira, J., 1993. Income distribution and macroeconomics. Review of Economic Studies 60, 35–52. Moav, O., 2002. Income distribution and macroeconomics: the persistence of inequality in a convex technology framework. Economics Letters 75, 187–192. Mushinski, D., Phillips, R.J., 2008. The role of Morris plan lending institutions in expanding consumer microcredit in the United States. In: Yago, G., Barth, J.R., Zeidman, B. (Eds.), Entrepreneurship in Emerging Domestic Markets: Barriers and Innovations. Springer, New York, pp. 121–139.
Contents lists available at ScienceDirect
Economics Letters journal homepage: www.elsevier.com/locate/ecolet
A credit market in early stages of economic development Hideki Nakamura ∗ , Tetsuya Nakajima Faculty of Economics, Osaka City University, 3-3-138 Sugimoto, Sumiyoshi, Osaka 558-8585, Japan
article
info
Article history: Received 16 March 2010 Received in revised form 9 February 2011 Accepted 18 March 2011 Available online 6 April 2011
abstract Even if the relatively rich and the poor are initially caught in a poverty trap, the relatively rich can escape poverty by receiving payments from the poor. Further accumulation of wealth by the rich allows the poor to escape poverty. © 2011 Elsevier B.V. All rights reserved.
JEL classification: D91 O15 O16 Keywords: Credit market Early stages of economic development Relatively rich Poor
1. Introduction Models explaining the persistence of poverty and inequality are typically based on a poverty-trap mechanism—individuals with wealth above some threshold converge to a high-income steady state and those below the threshold converge to a low-income steady state.1 However, one or two centuries ago, most of the ancestors of the relatively wealthy of today had little wealth. How did they escape poverty? By considering a trickle-down effect, Aghion and Bolton (1997) and Galor and Moav (2004) examined the dynamics of wealth distribution and showed that the poor could escape poverty when the rich continue accumulating their wealth. If not only the poor, but also the relatively rich are caught in a poverty trap, the economy cannot take off. Considering a credit market in early stages of economic development, this paper presents a theory that allows the relatively rich to accumulate wealth. Following Moav (2002) and Galor and Moav (2004), a poverty trap is driven by increasing saving rates in the form of transfers to the offspring. Poor families then converge to a corner solution of no transfer. However, the relatively rich – individuals with some wealth – accumulate more wealth by lending to the
∗
Corresponding author. Tel.: +81 6 6605 2271; fax: +81 6 6605 3066. E-mail address: [email protected] (H. Nakamura).
1 See, for example, Galor and Zeira (1993) which is pioneering work. 0165-1765/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.econlet.2011.03.011
poor, who pay a high interest rate as a result of limited supply of funds and a high demand. The relatively rich – once credit markets are well functioning – could accumulate wealth and escape a lowincome trap before the poor. Galor (2005) extensively surveys long-run transition models offering explanations for the takeoff from poverty to growth. This paper presents a theory that allows the relatively wealthier, even if poor according to modern standards, to accumulate wealth, and, thereby, change the properties of the dynamic system over time. Furthermore, as found in Demirgüc-Kunt and Levine (2009), financial improvement has a significant positive effect on the income of the poor because it boosts efficiency and the equality of opportunity. This paper highlights the importance of improvements in the credit market for the process of escaping poverty.2 First, it allows the relatively rich to become richer. Second, further improvements in credit markets and the accumulation of wealth by the rich allow for a reduction in the interest rate and a trickle-down effect. The poor then can escape poverty.3
2 In the US, many paid workers appeared in urban areas after the Civil War. Their wages were low and their employment was unstable, and they had almost no wealth. This prompted the emergence of credit markets. Furthermore, Morris Plan banks, the first one being established in 1910, contributed to increased credit availability by lowering interest rates. See Mushinski and Phillips (2008). 3 In some less developed countries, credit markets are not functioning well. For example, in Nepal, many households that have worked as bonded laborers in the
H. Nakamura, T. Nakajima / Economics Letters 112 (2011) 42–44
43
Taking (1) into account, the dynamics of investment in the case of (5) become eit = (1 − α)f (eit −1 ) − αθ .
(7)
Curve C in Fig. 1 depicts this equation. We consider the case that (7) has two stationary states: e∗ and e∗∗ . In Fig. 1, we assume that
(1 − α)f ′ e∗ < 1 and (1 − α)f ′ e∗∗ > 1. (8) Let us consider the case in which bp,−1 = 0 and 0 < br ,−1 <
e∗∗ . The bequest level of the poor remains zero. Because the initial wealth level of the relatively rich is below the threshold, it decreases and becomes zero after some time. Thus, economic development never occurs. 3. Model with a credit market Now, suppose that a credit market becomes available. Eqs. (5) and (6) are then modified as
Fig. 1. Dynamics of the investment level.
bit = (1 − α)Iit − αθ ,
2. Basic model with no credit market
bit = 0,
if (1 − α)Iit > αθ ,
(9)
if (1 − α)Iit ≤ αθ ,
(10)
Our model is a closed overlapping-generations economy. If parents decide to leave their bequests, their children receive these bequests in the first period. In the second period, they produce a single type of goods. We first consider the case in which there is no credit market. The output can be used for either consumption or bequest. A bequest is used for investment. Individuals are identical except for the bequests that are received from their parents. We assume that the initial numbers of rich and poor are ηL and (1 − η)L, respectively, where L represents the population of each generation. Each individual born in period t − 1 receives a bequest from their parent and spends it just for investment, because there is no credit market. That is, we have
where Irt and Ipt are the respective income levels of the relatively rich and the poor in period t. An individual determines the optimal investment level to maximize their income by borrowing or lending. The incomemaximization problem of the relatively rich, who are the lenders, is
eit −1 = bit −1 ,
brt = (1 − α) f (ert −1 ) + f ′ (ert −1 ) (brt −1 − ert −1 ) − αθ .
(1)
where i = r , p; brt −1 and bpt −1 are the respective bequest levels of the relatively rich and the poor that are received in period t − 1; and ert −1 and ept −1 are the respective investments in period t − 1. In period t, each individual produces goods subject to the following production function yit = f (eit −1 ),
(2)
where yrt and ypt are the respective levels of outputs in period t. We assume that f ′ (eit −1 ) > 0 and f ′′ (eit −1 ) < 0. We also assume that f (0) > 0. We assume the altruistic bequest motive, i.e., the ‘joy of giving’. A zero bequest is allowed because individuals can live with no investment. The utility maximization of an individual born in period t − 1 is as follows max α ln cit + (1 − α) ln(bit + θ ), cit ,bit
0 < α < 1, θ > 0,
(3)
max Irt = f (ert −1 ) + rt (brt −1 − ert −1 ) ,
(11)
ert −1
where rt is the gross interest rate in the credit market. The investment level is determined as f ′ (ert −1 ) = rt .
(12)
Eq. (9) is rewritten as
(13)
On the other hand, the poor, who have no wealth, become the borrowers. Because the lenders to individuals have positive costs of keeping track of each borrower, the individual must borrow at a rate higher than rt . They solve max Ipt = f ept −1 − δ rt ept −1 ,
ept −1
δ ≥ 1.
(14)
The investment level of the poor is lower than that of the relatively rich because of an imperfect credit market f ′ ept −1 = δ rt .
(15)
Given the interest rate, the cost of keeping track increases with an increase in the investment level. Borrowers must pay higher interests as δ rises. 4. Can the credit market help the relatively rich escape poverty? The credit market equilibrium is given by
s.t. cit + bit = f (eit −1 ),
(4)
η(brt − ert ) = (1 − η)ept .
(16)
where crt and cpt are the respective consumption levels. We assume that (1 − α)f (0) < αθ . The first-order conditions yield
The investments of the relatively rich and the poor are financed by the wealth of the relatively rich only. We specify the production function as
bit = (1 − α)f (eit −1 ) − αθ ,
(5)
f (eit ) = a0 + a1 eit ,
(6)
Given (17), the investment level of the poor is proportionate to that of the relatively rich
bit = 0,
if (1 − α)f (eit −1 ) > αθ ,
if (1 − α)f (eit −1 ) ≤ αθ .
When the income level is low, the bequest level becomes zero because of the shift parameter, θ .
γ
ept = gert ,
(17)
(18) −1/(1−γ )
tea industry manage their living expenditure by further borrowing at high interest rates.
a0 , a1 > 0, 0 < γ < 1.
where g ≡ δ ≤ 1. While an increase in g implies an improvement in credit availability, the difference in investment levels between the relatively rich and the poor becomes small. Using (13), (16) and (18),
44
H. Nakamura, T. Nakajima / Economics Letters 112 (2011) 42–44
the dynamics of the investment level can be represented as ert =
1−α
ηf (ert −1 ) + (1 − η) gf ′ (ert −1 ) ert −1
η + (1 − η) g ηαθ − . η + (1 − η) g
(19)
The relatively rich obtain not only income from their own investment, but also the interest payments from the poor that increase with an improvement in the degree of credit availability. Eq. (19) is drawn as curve D in Fig. 1. We assume that curve D is strictly concave. In Fig. 1, eD indicates a stable stationary state, and eDD is an unstable stationary state. Curve D approaches curve C as g decreases.4 We define eˆ as an investment that satisfies
(1 − α) f eˆ − f ′ eˆ eˆ − αθ = 0.
(20)
Curves C and D intersect at the point where ert −1 = eˆ regardless of the degree of credit market imperfection. How does an improvement in credit availability affect the investment levels of stationary states? The comparative statistics with respect to g are
where
{(1−α)[f (e)−f ′ (e)e]−αθ } = − 1−[(1−α)/(η+(1−η)g )]{[η+(1−η)g]f ′ (e)+(1−η)gf ′′ (e)e} . η(1−η)/(η+(1−η)g )2
An improvement in credit availability decreases the investment levels of the relatively rich on the stationary states. When credit availability is improved, the wealth of the relatively rich is more efficiently allocated between the relatively rich and the poor. The concavity of the production technology implies that the investment level of the relatively rich decreases while the level for the poor increases. The income and wealth levels of the relatively rich on the high-level stationary state increase because of an increase in the interest payments from the poor to the relatively rich. The income level of the poor on the stationary state also increases because of an increase in their investments. It is possible for the relatively rich to escape poverty because of the existence of a credit market. The relatively rich can accumulate their wealth if the initial investment level is greater than eDD , which is lower than the threshold e∗∗ , i.e., eDD < er ,−1 =
η br ,−1 < br ,−1 < e∗∗ . η + (1 − η) g
5. Conclusion We considered an economy in the early stages of economic development, where not only the poor but also the relatively rich are initially caught in poverty. We found that a credit market potentially has the power to help the relatively rich accumulate their wealth before the poor.
∂ eD ∂ eDD < 0 and < 0, ∂g ∂g ∂e ∂g
In the early stages of economic development, many individuals have no wealth to invest. A few individuals who have small wealth exist. However, investing all their wealth in their own production cannot yield enough income to escape the low-income trap, because production is subject to diminishing returns. If a credit market is available, the relatively rich lend a part of their wealth to the poor. The relatively rich can receive a large amount of interest payments from the poor as a result of limited supply of funds and a high demand. The relatively rich can then escape the trap by accumulating their wealth. Thus, a credit market with a sufficient degree of credit availability can make an economy take off because of the resource-allocation effect and the incomedistribution effect. In addition, we examine whether the poor can escape poverty. If the investment level of the poor exceeds eˆ , defined in (20), until the investment level of the relatively rich converges to eD , the poor can start accumulating their own wealth. There exists g˜ that satisfies g˜ eD = eˆ . Therefore, there exists a trickle-down effect if a credit market is sufficiently improved.
(21)
We also obtain ∂ br /∂ g < 0 at e = eDD , where br = eDD [η/(η + (1 − η)g )]−1 . Thus, it becomes easy for the relatively rich to escape poverty by improving credit availability.5 Here, we have the following proposition. Proposition 1. Given (21), a credit market can help the relatively rich, who are caught in a poverty trap, to escape the trap.
4 The proof, which shows that eDD < e∗∗ < eD < e∗ , is available on request. 5 Furthermore, (21) can hold when the ratio of the relatively rich to the total population is small.
Acknowledgments We would like to thank Yoshihiko Seoka, Mario Oshima, and the seminar participants of Osaka University of Economics for their comments. A referee provided helpful suggestions to improve the paper. References Aghion, P., Bolton, P., 1997. A theory of trickle-down growth and development. Review of Economic Studies 64, 151–172. Demirgüc-Kunt, A., Levine, R., 2009. Finance and inequality: theory and evidence. NBER Working Paper Series 15275. Galor, O., 2005. From stagnation to growth: unified growth theory. In: Aghion, P., Durlauf, S. (Eds.), Handbook of Economic Growth, vol. 1A. Amsterdam, NorthHolland, pp. 171–293. Galor, O., Moav, O., 2004. From physical to human capital accumulation: inequality and the process of development. Review of Economic Studies 71, 1001–1026. Galor, O., Zeira, J., 1993. Income distribution and macroeconomics. Review of Economic Studies 60, 35–52. Moav, O., 2002. Income distribution and macroeconomics: the persistence of inequality in a convex technology framework. Economics Letters 75, 187–192. Mushinski, D., Phillips, R.J., 2008. The role of Morris plan lending institutions in expanding consumer microcredit in the United States. In: Yago, G., Barth, J.R., Zeidman, B. (Eds.), Entrepreneurship in Emerging Domestic Markets: Barriers and Innovations. Springer, New York, pp. 121–139.