As Bewley reminds us, labor market conditions ‘external’ to the firm play a crucial role in the
determination of wages. Since wages determine the return on human capital, external conditions
are also crucial for the incentive of workers to accumulate human capital. In this paper we show
that ‘external competitive’ conditions are important if contracts are not enforceable for both
employers and employees (double-sided limited commitment).
It is not difficult to see why in reality contracts are not fully enforceable for both employers
and employees. On the one hand, without slavery or serfdom, workers are free to move to obtain
the best return on their human capital. Together with the limited observability of workers’ effort,
this reduces the enforcement of contracts for workers. On the other hand, employers could renege
on the compensation promised to workers for their effort to accumulate human capital. Under
these circumstances, workers would be discouraged from accumulating skills unless their human
capital can be easily redeployed outside the firm. Since this depends on the competitiveness of
the labor market, competition plays a crucial role in affecting the accumulation of human capital.
This mechanism, which was already informally described by Adam Smith but until now has not
been extensively explored, is the central focus of this paper.
We use a dynamic general equilibrium model where contracts are not enforceable, neither
for workers nor for firms (employers). We show that the way limited enforcement affects the
accumulation of ‘general’ human capital depends on barriers to the mobility of skilled labor. In
particular, high barriers discourage the accumulation of human capital while low barriers have
a stimulating effect, both in new and incumbent firms. As a result, differences in ‘barriers to
competition’ translate into significant differences in incomes and welfare across economies.
While this result vindicates the common wisdom that ‘barriers to the external conditions of
the firm’ are ‘barriers to riches’, it also shows that this effect can be greatly amplified when
contracts are not enforceable for either party. In fact, a key finding of the paper is that the limited
enforcement of contracts matters for the accumulation of human capital only when the limited
commitment is double-sided, that is, when both sides can renege on the contract. Instead, with
the commitment of at least one party, barriers to the mobility of skilled labor have limited effects
on the accumulation of human capital and aggregate output.
The central mechanism through which barriers to the mobility of skilled labor affect the accu-
mulation of ‘general’ human capital is by reducing the outside opportunities of skilled workers,
that is, the value of redeploying their skills outside the firm. With a lower outside value, workers
do not have a credible mechanism for punishing the firm for reneging on the promised compen-
sation. Anticipating this, the worker makes less effort to acquire skills, and therefore, there is
a typical time-inconsistency problem. It is important to emphasize that the analysis focuses on
the type of human capital that is not specific to an individual firm, and therefore, it can be trans-
ferred to other firms. Otherwise, the outside value of the worker would be independent of human
capital.
Why do we need double-sided limited commitment? If the worker could commit to staying
with the firm and providing effort (one-sided commitment from the worker), the contractual
friction associated with the limited commitment of the firm could be resolved by paying the worker in advance. However, without the commitment from the worker, advance payments or
ex-post payments conditional on the productivity realized are not incentive-compatible. This is
because the worker can simply exert less effort and quit. On the other hand, if the investor could
commit, the promised payments would not be reneged on ex-post. Thus, it becomes feasible
to implement an incentive-compatible wage mechanism to extract the right level of effort and
prevent the worker from quitting. For example, in line with existing Contract Theory, one could
solve the commitment problem with an output-sharing agreement or by transferring the total or
partial ownership of assets to the workers (e.g. [14]). But with a two-sided lack of commitment,
such arrangements are still open to unverifiable
de facto
renegotiations or skimming.
The only way to make the contract free from renegotiation is by choosing a level of invest-
ment in human capital that makes the
ex-ante
payment promised by the firm exactly equal to
the
ex-post
outside value of the worker, making renegotiation superfluous. In our economy, the
best outside value for the worker is the one received by entering into a contractual arrangement
with a new firm. Therefore, a credible investment policy for an incumbent firm is to mimic the
investment decision of a new firm. However, when the investment cost of new firms is high (that
is, there are high barriers) their investment is low, implying that the investment by incumbent
firms is also low. In contrast, with full or one-sided commitment, incumbent firms do not have
to mimic the investment decisions of new firms. In summary, in an economy with full or one-
sided commitment, barriers to competition affect only the human capital accumulation of new
firms. However, in an economy with two-sided limited commitment, barriers affect the invest-
ment decisions of
all
firms. In particular, in economies where in equilibrium there is no exit or
entry, changing the degree of competitiveness may have no effect if there is one-sided limited
commitment. If there is two-sided limited commitment instead, the effects could be substantial.
Our results are first illustrated with a simple two-stage model which is then extended to a
dynamic infinite horizon set-up. The parameterization of the infinite horizon model allows us to
quantify the ability of one particular barrier—
start-up costs
—to account for different levels of
human capital accumulation, as well as cross-country income differences. The baseline model
can account, roughly, for half of the cross-country income gaps with the US. Even though this
number should be taken with caution, given the simplicity of the model, it shows that this mech-
anism can be quantitatively important, bringing a new perspective on the role of competition as
a factor of growth.
In our benchmark model, ‘barriers to the external conditions of the firm’ are given by the cost
of starting a new firm (which offers the best outside option to the skilled worker). However, there
exist many other barriers that could generate similar qualitative results. For example, we show
that
covenants
(preventing a skilled worker from working for a different employer in the same
industry for a period of time) or strong enforcement of Intellectual Property Rights can depress
human capital accumulation and possibly explain regional and national income differences.
The results of this paper can also be interpreted as saying that
barriers
to competition de-
termine cross-country positions relative to the ‘technology possibility frontier’, without empha-
sizing a distinction between innovation and technology adoption. This is consistent with the
idea that even the implementation of known technologies requires appropriate human capital. In
contrast, Acemoglu et al. [4] develop a theory where the ‘distance to the frontier’ determines
a country’s comparative advantage in innovation vs. adoption. While in their theory the cost of
barriers depends on the position of a country relative to the frontier, in our framework it is the
barriers that determine the position of a country in relation to the frontier. The causality effect is
reversed and the policy implications are different. They show that a lack of pro-competition poli-
cies becomes more costly as countries approach the world technology frontier, while our theoryimplies that a lack of pro-competition policies can determine a country’s position away from the
frontier.
Amaral and Quintin [5] also show that limited enforcement of contracts can have a large
impact on equilibrium output because it reduces the capital directed to the production sector
and the employment of efficient technologies. The main mechanism relies on the accumulation
of ‘physical’ capital when contracts are not enforceable for entrepreneurs (one-sided limited
enforcement). The accumulation of physical capital is also the key mechanism in [9] and [13].
Our paper, instead, focuses on the accumulation of ‘human’ capital when there is double-sided
limited enforcement and emphasizes the central role played by ‘competition’ for human capital.
In contrast to these papers, we show that more enforcement is not necessarily welfare enhancing
when there are externalities in human capital accumulation.
This paper relates to multiple strands of literature. In addition to the studies already cited, at
least three more should be mentioned. First, it relates to the extensive literature on growth the-
ories based on endogenous human capital accumulation (e.g. [18,23]). In these theories, human
capital is rewarded through the usual channel of higher competitive wages. We take a closer look
at this channel, showing how it can be affected by the interplay between competition and commit-
ment in the skilled labor market. Second, our paper relates to the labor literature that studies the
accumulation of skills either within the firm (e.g. [2]) or before workers and firms are matched
and wages are negotiated. In most of this literature (e.g. [1,3]), higher outside values worsen the
hold-up problem leading to lower accumulation of skills. In our framework, stronger competi-
tion results in higher outside values.
Third, the paper relates to the literature that emphasizes
the role of ’barriers to riches’ in explaining income differences [22,24]. We emphasize the im-
portance of barriers to labor mobility and how the effects of these barriers depend on contractual
commitment features.
We have developed a theory in which
barriers
to the mobility of skilled workers affect the
accumulation of human capital or knowledge, and therefore, the level of income. The theory
does not simply say that competition enhances income. First, it emphasizes that some forms of
limited enforcement are intrinsic to competitive labor markets, where wages are determined by
supply and demand conditions. Second, it shows how different forms of contract enforcement
affect the relation between competition, accumulation of human capital and economic develop-
ment. In particular, when both investors and workers cannot commit to long-term contracts, the
accumulation of human capital is determined by those firms that value human capital the most;
in our benchmark model they are start-up firms.
In this way our theory captures Bewley’s view that “
Wages rise quickly and sometimes dramat-
ically in response to increases in the market demand for certain type of labor, but these increases
are a reaction to competition from other firms, not to internal pressure from employees
”, Bewley
[6, p. 407]. Our contribution is to show that ‘competition from other firms’ is also important
for the accumulation of human capital. In particular, high levels of human capital are associated
with low barriers to the mobility of knowledge because lower barriers increase ‘competition from
other firms’.
Using a semi-endogenous growth model, we have shown that
barriers to business start-up
have the potential to explain significant cross-country income differences. It may seem that
accounting for 50% of the cross-country income gaps with the US overestimates the real contri-
bution of our mechanism since the model is silent about many other features that are important for developing economies. Nevertheless, the fact that investment in human capital—even at the school level—is determined by expectations of future income returns, which in turn are affected
by competitive and contractual conditions, seems a powerful mechanism.We have also shown that other
barriers to knowledge mobility
, such as strict enforcement of
Covenants or Intellectual Property Rights, can have similar effects. Although we have modeled
‘on-the-job human capital accumulation’ and abstracted from skill-specific jobs, it should be
clear that the mechanism described here is also relevant for ‘before-the-job human capital ac-
cumulation’. In particular, the returns to education or to acquiring a specific mix of general and
specific skills, depend on expected life-time returns. Our model shows that these returns depend crucially on the interplay between competition for skills in the labor market and the degree of
commitment and enforcement.