Job creators, job creation, and the tax code

• Date: 01 June 2017
• Published date: 01 June 2017
• DOI: http://doi.org/10.1111/jpet.12230
• Author: Adrian Masters

Received: 12 September 2016 Accepted: 23 September 2016

DOI: 10.1111/jpet.12230

ARTICLE

Job creators, job creation, and the tax code

Adrian Masters

Universityat Albany

AdrianMasters, Department of Economics,

Universityat Albany, SUNY 1400 Washing-

tonAvenue Albany,NY 12222 (amasters@

albany.edu).

This paper considers the role of the tax code in determining

income dispersion and vacancy creation. A “span-of-control” model

is embedded into a search and matching environment. A cut to the

tax on profits in isolation improves job creation and reduces before-

tax income inequality. The impact of a budget-balancing increase in

the wage tax depends on the bargaining power of firms. When it is

high, firms pick up the lion’s share of the tax burden. The tax acts like

a barrier to entry: it benefits large firms at the expense of marginal

ones. Net effects are an increase in unemployment and before-tax

income dispersion. Low firm bargaining power means workers pick

up more of the tax burden. It acts like a subsidy to entrepreneur-

ship reinforcing the impact of the profit tax reduction. Taxes on the

returns to capital leave everyoneworse off.

“Taxhikes destroy jobs—especially an increase on the magnitude set for January 1st.” John Boehner’s address to

Peter G. Peterson Foundation2012 Fiscal Summit.

“We need serious tax reform to make the tax code fairer and simpler. The most profitable corporations should

haveto pay their fair share …those who already have made it big have a responsibility to pay a little bit forward—

so the next kid coming along has a chance to makeit too.” Elizabeth Warren for Senate, Issues website.

1INTRODUCTION

Depending on your political position, lowering taxes on profits either stimulates vacancy creation or exacerbates

income inequality. These assertions are not mutually exclusive. The goal of this paper is to provide a parsimonious

model of entrepreneurship and job creation to understand when and how either or both of them can be true.

The model embeds ex ante heterogeneity in embodied entrepreneurial ability into a search and matching frame-

work. Individuals choose to either establish a firm or remain as a worker.Individual firm owners receive rents in the

form of profits above those accruing to the marginal entrepreneur. Taken in isolation, a cut in the tax rate on profits

increases the return to entrepreneurship—more firms are created and unemployment falls. Because there are more

firms to share a reduced population of workers, before-tax profits decline. In the longer term, however, the govern-

ment must balance its budget. This is achieved in the model by raising the tax on wages, the impact of which depends

on the bargaining power of the firms. When it is high, firms pick up the lion’s share of the tax burden in the form of a

wage increase. The tax acts like a barrier to entry which benefits large firms at the expense of marginal ones. The net

Journal of Public Economic Theory 2017; 19: 674–691 wileyonlinelibrary.com/journal/jpet c

2017 Wiley Periodicals,Inc. 674

MASTERS 675

effects are an increase in unemployment and before-tax income dispersion. Low firm bargaining power means work-

ers pick up more of the increased wage-tax burden and it acts like a subsidy to entrepreneurship. This reinforces the

impact of the profit-tax reduction.

The framework introduces the Lucas (1978) “span-of-control” model into a Diamond-Mortensen-Pissarides (DMP)

style search and matching environment. As in Lucas’s model, anyone can choose to be either a worker or an

entrepreneur/firm owner. There is ex ante heterogeneity in terms of individuals’ entrepreneurial abilities but as a

worker everyoneis equally productive. Consequently, there is a threshold level of entrepreneurial ability above which

individuals choose to establish a firm and below which they do not. Firms hire workers through a DMP environment

that is governed by a matching function.

In a departure from the usual DMP set-up, I allow firms to hire multiple workers.1Firms effectively operate a

standard neoclassical production function with constant returns to scale. Factors of production are capital and labor.

Although there are diminishing returns to each factor,when a worker is hired the firm acquires the appropriate amount

of additional capital through a competitive market. Output is therefore linear in the number of workers. Total factor

productivity (TFP) of the firm is specified by the owner’s entrepreneurial ability level.Firms are continually in the labor

market and they hire whomever they meet. What prevents them from growing without bound is that, as is common

in the DMP framework, there is an exogenous rate of separation. Consequently, the number of workers at any firm

is (ergodically) Poisson distributed. With a constant rate of separation, the parameter of the Poisson distribution is

determined by the matching rate.

Workers can observe the firm’s TFP and they direct their search accordingly.Wage formation is by Nash bargain-

ing.2Because the effective marginal product of the workers is constant, the wage depends on the firm’s TFP but not

on the current number of employees. Higher TFP means higher marginal productivity and higher wages. More work-

ers, therefore, seek employment at the higher TFP firms and such firms experiencea higher matching rate. So, not only

do the highest ability entrepreneurs receive higher rents from each employee, they also (in expectation)have more of

them.

In this context,the paper asks how the tax code affects vacancy creation and the relative earnings of the high-ability

entrepreneurs. Like Lucas (1978), the model lacks a general pyramidal organizational structure with severallayers of

management in each firm. The model cannot, therefore, empirically match both the firms’ size and income distributions.

Here, the focus is on the income distribution. The central exercise is then to set thetax on profits to 15% and set the

wage-tax so as to generate revenues at 18.6% of GDP. Then, a common tax rate is sought to generate the same tax

revenue. Toseparately identify the effects of the tax on profits from the tax on wages, I report the incremental results

as the economy movesfrom the equal to the unequal tax regime.

Central to the literature onthe calibration of DMP models is the value of the bargaining power parameter. The cru-

cial point is that setting the bargaining power of the firms equal to the elasticity of the matching function with respect

to vacancies (the Hosios Rule) implements the constrainedefficient allocation. Estimates of the elasticity of the match-

ing function (see Petrongolo and Pissarides, 2001) range between 0.3 and 0.7 with most people using 0.5. Meanwhile,

empirical studies of micro data imply that the bargaining power of the firms exceeds 0.95 (see Card, Devicienti, and

Maida, 2014). The problem is that such a bargaining power puts the laissez faire economy so far from efficiency as to

stretch credulity. Consequently,I provide two separate parameterizations that serve to demonstrate the workings of

the model. The first is based on the empirical measures of the bargaining power (I use 0.96) and the second is based on

the Hosios (1990) result (I use 0.5).

In both cases, we see that the reduction of profit taxes in isolation increases the return to entrepreneurship, job

creation improves, before-tax profits decline, and after-tax profits rise. The impact of the budget-balancing increase in

the wage-tax depends on the parameterization. With the bargaining power of firms set to 0.96, the increased wage-tax

1Otherpapers that incorporate multiple-worker firms in a matching environment include Burdett and Mortensen (1998), Acemoglu and Hawkins (2014), and

Shimer(2006).

2Some researchers have criticized directed search because it has historically relied on both posted wages and full commitment to those wages while there

is little evidence that either aspect of the wage formation process is true (see Masters, 2011; Menzio, 2007). Here, there is no commitment issue and in

equilibrium,at least, entrepreneurial ability will be evident by the behavior of the firm.