Local Multipliers - Econometrics Laboratory, UC Berkeley

1 American Economic Review: Papers & Proceedings 100 (May 2010): 1 7. Local Multipliers By Enrico Moretti*. Every time a Local economy generates a new The magnitude of Local Multipliers is impor- job by attracting a new business, additional jobs tant for regional economic development policies. might also be created, mainly through increased State and Local governments spend consider- demand for Local goods and services. This posi- able amounts of taxpayer money on incentives tive effect on employment is partially offset by to attract new businesses to their jurisdictions. general equilibrium effects induced by changes Such location-based incentives are pervasive in in Local wages and prices of Local services. In manufacturing. However, the efficiency of these this paper, I estimate the long-term employment policies and their actual effects on employment multiplier at the Local level. Specifically, I quan- are not fully understood, because there is little tify the long-term change in the number of jobs systematic evidence on the effects of success- in a city's tradable and nontradable sectors gen- fully attracting a new firm on other parts of the erated by an exogenous increase in the number Local economy.

2 The estimates in this paper help of jobs in the tradable sector, allowing for the inform this endogenous reallocation of factors and adjust- Moreover, assuming that the objective of Local ment of prices. economic development policies is to maximize I find that for each additional job in manu- Local employment, it is important to know where facturing in a given city, jobs are created in subsidies should be directed. The multiplier is the nontradable sector in the same city. As the likely to vary across industries and skill groups. number of workers and the equilibrium wage There is little existing evidence on which indus- increase in a city, the demand for Local goods and tries and skill groups have the largest multiplier services increases. This effect is significantly and therefore generate the largest number of larger for skilled jobs, because they command additional jobs. My estimates shed some light higher earnings. Adding one additional skilled on this question. job in the tradable sector generates jobs in It should be noted, however, that the presence Local goods and services.

3 The corresponding of large Multipliers is not, in itself, a market figure for unskilled jobs is one. The multiplier failure and therefore does not necessarily jus- also varies across industries. Industry-specific tify government intervention. Local subsidies Multipliers indicate that high tech industries may be efficiency enhancing in the presence of have the largest multiplier . agglomeration externalities. However, a multi- A simple framework suggests that the Local plier larger than one does not necessarily imply multiplier for the tradable sector should be the existence of agglomeration economies. For smaller than the one for the nontradable sector, example, the multiplier effect that operates and possibly even negative. This is because the through increases in the product demand for increase in labor costs generated by the initial Local goods and services is a pecuniary exter- labor demand shock hurts Local producers of nality and does not constitute a market failure. tradables. This negative effect may be in part On the other hand, the finding of a nonnegative offset by agglomeration externalities, if they employment effect for tradables is consistent exist, and an increase in the demand for inter- with (although not proof of) the existence of mediate inputs, if supply chains are localized.

4 Agglomeration economies. Empirically, I find that adding one additional The magnitude of Local Multipliers may also job in one part of the tradable sector has no sig- be relevant for the literature on nationwide mul- nificant effect on employment in other parts of tipliers. The exact magnitude of Multipliers is the tradable sector. * Department of Economics, UC Berkeley , CA 1. See also Greenstone and Moretti (2004); and 94720 3880 (e-mail: Greenstone, Hornbeck, and Moretti (forthcoming). 1. 1 3/31/10 11:57 AM. 2 AEA PAPERS AND PROCEEDINGS MAY 2010. a crucial element in formulating countercycli- results in an increase in the Local demand for cal stimulus policies. For example, the Obama nontradables z1, z2, , z M. Employment in indus- administration's ex ante estimates of the effect of tries like restaurants, real estate, cleaning ser- the 2009 American Recovery and Reinvestment vices, legal services, construction, medical Plan depended crucially on the magnitude of the services, retail, personal services, etc.)

5 Grows multiplier used in the simulations (Romer and because the city has more workers and wages Bernstein 2009). Existing estimates of multipli- are higher. These new jobs are split between ers based on national time-series hinge upon existing residents and new residents who move very strong identifying assumptions. My esti- from somewhere else, depending on the degree mates of Local Multipliers provide bounds for of geographical mobility. The magnitude of the national Multipliers . multiplier effect depends on three factors. First, it depends on consumer preferences for nontrad- I. Conceptual Framework ables and the technology in the nontradable sec- tor. More labor intensive technologies result in a Assume that each city is a competitive larger multiplier . Second, it depends on the type economy that uses labor to produce a vector of new jobs in the tradable sector. Skilled jobs of nationally traded goods, x1, x2, x3, , xK should have a larger multiplier than unskilled whose price is set and a vector of nontraded jobs, because they pay higher earnings and goods, z1, z2, , z M whose price is determined therefore are likely to generate a larger increase locally.

6 Labor is mobile across sectors within in the demand for Local services. a city so that marginal product and wages are Third, there are offsetting general equilib- equalized within a city. Local labor supply rium effects on wages and prices, which ulti- is upward sloping, and its slope depends on mately depend on the elasticities of Local labor the distribution of residents' tastes for leisure and housing supply. The citywide increase in and the degree of labor mobility across cities. labor costs generated by the shock to x1 causes Higher geographical mobility implies a higher a decline in the supply of Local This elasticity of labor supply. In the extreme, perfect decline partially but not fully undoes the mobility would imply an infinitely elastic Local effect of the increase in demand for Local ser- supply of labor. Local housing supply is upward vices. Effectively, the addition of jobs in x1 par- sloping, and its slope depends on geography and tially crowds out jobs in other industries.

7 If labor land use regulations. supply is locally very elastic, this crowding out Consider the case of a permanent increase is more limited and the increase in labor costs is in labor demand in tradable industry x1 in city small, making the multiplier c. This increase may be due to the successful attraction of a new firm (see for examples the multiplier for the Tradable Sector. The cases documented in Greenstone and Moretti shock to industry x1 may also affect employment 2004) or an increase in the product demand in tradable industries x2, x3, , xK although the faced by existing firms. The direct effect of this direction of the effect is a priori unclear. This shock is an increase in employment in industry effect is governed by three different forces. x1. But this shock to sector x1 may also affect First, and most important, the citywide increase Local employment in the rest of the tradable sec- in labor costs hurts employment in x2, x3, , xK. tor x2, x3, , xK and in the nontradable sector.

8 Because these are tradable industries, the The shock is also likely to have general equi- increase in production costs lowers their compet- librium effects on Local prices: the wage of all itiveness. Unlike the case of nontradable goods, workers in the city increases (unless Local labor supply is infinitely elastic) and the cost of hous- 2. ing also increases (unless housing supply is infi- This decline is further exacerbated by the increase in the cost of land caused by the increase in population. nitely elastic). 3. In the extreme case where Local labor supply elasticity is infinite, nominal wages in the city do increase, but only to multiplier for the Nontradable Sector. The compensate workers for the higher cost of housing, so that real wages remain constant. In this case, the decline in the city budget constraint increases, both because supply of nontradables is limited, and the increase in the there are more Local jobs and because wages are demand for Local services is driven only by the increase in higher.

9 The increase in the city budget constraint number of workers in the city. 2 3/31/10 11:57 AM. VOL. 100 NO. 2 Local Multipliers 3. the price of tradable goods is set on the national bound for the national multiplier . By definition, market and cannot adjust to Local economic con- the market for tradables is national, and much of ditions. In the long run, some of the production the additional Local demand is likely to benefit in these industries is likely to be shifted to dif- other cities. Additionally, the negative effects of ferent cities. Second, the increase in production higher labor costs are more significant locally of x1 may increase the Local demand for inter- than nationally. mediate goods and services. This effect depends on the geography of the industry supply chain. II. Empirical Estimates While many industries are geographically clus- tered, the magnitude of this effect is likely to Using data from the 1980, 1990, and 2000. be quantitatively limited if the market for x2, Census of Population, I estimate variants of the x3, , xK is truly Third, if agglomera- following models: tion economies are important, the increase in production in sector x1 may result in more Local (1) N c NT.

10 T = + . Nc Tt + dt + ct agglomeration (see, for example, Greenstone, Hornbeck, and Moretti forthcoming). 5 (2) Nct T1 = + Nct T2 + dt + ct State and Local governments spend consider- able public resources to finance regional eco- where N c Tt and N c NT. t are the change over time nomic development policies. To estimate the in the log number of jobs in city c in the trad- economic impact of these policies, state and able and nontradable sector, respectively; Nct T1 . Local governments typically use estimates of is the change in the log number of jobs in a ran- Local Multipliers based on Local Input-Output domly selected part of the tradable sector; and tables. This simple framework shows that Input- Nct T2 is the change in the log number of jobs Output tables are unlikely to produce meaning- in the rest of the tradable sector. The sample ful estimates of Local Multipliers . First, they includes two observations per city, correspond- completely miss the employment effect for ing to the periods 1980 1990 and 1990 2000.