Evaluating the effects of local content measures in a CGE model: Eliminating the US Buy America(n) programs
Introduction
Governments in many countries implement local-content policies in government procurement. The U.S. is no exception. Since the Buy American Act of 1933 and earlier,1 the U.S. federal government has tried to protect manufacturing industries such as steel by channeling its purchases of goods and construction projects towards U.S. suppliers. This includes the federal government's own direct purchases, as well as purchases by its instrumentalities such as Amtrak, together with purchases by state governments using federal funds. While the Buy American Act operates on direct purchases by government agencies, more recent Buy America schemes operate on indirect purchases by constraining the input purchases made by firms which produce goods sold to government agencies. Through Buy America schemes, the U.S. government has tried to reach beyond the nationality of its direct suppliers through to the national origin of the inputs that they use. For instance, U.S. contractors supplying construction projects to the public sector financed under the American Recovery and Reinvestment Act (ARRA, the Obama stimulus package of 2009) were obliged to use U.S.-produced steel and other manufactured inputs. As described in Koehl and Masini (2017), when the federal government purchases an oven for a military mess, they inquire into the national origin of the component parts, such as the oven door handle.
We use the term “Buy America(n)” to refer generally to schemes which favor local content in both direct and indirect public-sector purchases. Implementation of Buy America(n) is governed by highly detailed regulations. For example, the Secretary of Defense is required “to encourage increased domestic breeding while ensuring that military working dogs are procured as efficiently as possible and at best value to the government” (see Manuel et al., 2016). Regulations at this level of detail are subject to expensive legal interpretations and litigation (see Koehl and Masini, 2017). To us, they seem a fruitful area for legislators interested in finding regulations that can be scrapped when trying to comply with the spirit of President Trump's demand for the elimination of two regulations whenever a new one is introduced (see Mufson, 2017).
Drawing on the Government Accountability Office (GAO), Hufbauer et al. (2013) summarizes the general aims of Buy America(n) as:
- •
boosting domestic employment and economic growth through infrastructure spending;
- •
protecting against unfair competition from foreign firms that receive subsidies from their governments; and
- •
strengthening national security by promoting the iron and steel industries.
In this paper, we abstract from the minutia of Buy America(n). We use USAGE, a detailed computable general equilibrium (CGE) model, to throw light on the issue of whether such schemes could ever be expected to deliver on their objectives. Hufbauer et al. (2013) list various obvious problems with Buy America(n) including higher costs to government, reduced bidding competition, project delays while compliance is being worked out, and potential international retaliation.2 But we assume that Buy America(n) works in a comprehensive transparent way and is tolerated by foreigners. We show that even under these favorable conditions such schemes are likely to be counter-productive.
The paper is organized as follows. Section 2 provides a brief introduction to USAGE. Section 3 explains our methodology, including how we represent Buy America(n) in USAGE. 4 Simulating the elimination of Buy America(n) in the USAGE model: macro assumptions and results, 5 Effects of eliminating Buy America(n) on U.S. output, 6 Effects of eliminating Buy America(n) on employment in states and congressional districts give USAGE results at the macro, industry and regional levels for an illustrative simulation of the effects of eliminating Buy America(n). Concluding remarks are in Section 7.
Section snippets
Why a model and why the USAGE model?
We start by looking at the U.S. economy under the assumption that comprehensive Buy America(n) policies are in place. Then we work out the effects of the policies by calculating how the economy would be affected if they were removed.
The only feasible way of doing this is to apply a CGE model. These models link all the various parts of the economy so that we can trace out how a switch towards imports in government-financed projects affects:
- •
the balance of payment and the exchange rate;
- •
output and
Representing Buy America(n) in the USAGE model
Buy America(n) provisions operate on both direct purchases by government agencies and indirect purchases. Direct refers to purchases made by government agencies while indirect refers to purchases made by firms in creating goods sold to government agencies. For example, the government directly buys a mile of paved road and indirectly buys asphalt used by the contractors who supply the paved road.
In practice, Buy America provisions (ie: those concerned with indirect inputs to government
Simulating the elimination of Buy America(n) in the USAGE model: macro assumptions and results
As explained in the previous section, we view the impact of the elimination of Buy America(n) as an array of technical changes. These cause the substitution of imported inputs for domestic inputs and introduce efficiency gains associated with freeing industry choices between domestic and imported inputs.
Before we can use USAGE to simulate the effects of any set of shocks (in this case technology shocks) we must set the closure. This refers to the macroeconomic assumptions. The main closure
Effects of eliminating Buy America(n) on U.S. output
Table 5.1 shows the effects of eliminating Buy America(n) programs on U.S. outputs of selected commodities. A priori we expected a large negative effect for U.S. output of any commodity c for which the technical change shocks TC(c,dom,j) calculated in Section 3 are large negative numbers for industries j that are important customers for domestically produced commodity c. More formally we expected output effects across all commodities c to be correlated with TCave(c) defined by:
Effects of eliminating Buy America(n) on employment in states and congressional districts
Table 6.1 shows job effects in the 51 states (includes the District of Columbia) from eliminating Buy America(n) calculated by the top-down method outlined in Section 2. It is the nature of trade policies to reallocate employment between a country's regions. This is because trade policies reallocate resources between a country's industries and for many industries, especially those producing traded goods, there is strong regional specialization. Regions specializing in industries that gain from
Concluding remarks
Like all local content programs for government procurement, Buy America(n) schemes are seductively attractive to politicians and the public more generally. What could possibly be wrong with channeling public expenditures towards U.S. producers? Economic modelling helps us to understand what is wrong.
Buy America(n) increases the costs to the U.S. government of infrastructure projects. With binding budget constraints, this means that governments can undertake a lower volume of projects than would
Acknowledgements
We thank the Canadian Embassy in Washington DC for financial support. We thank Niall Cronin and Carrie Goodge O’Brien from the Embassy and Shenjie Chen from Global Affairs Canada for valuable comments on earlier drafts. The analysis and conclusions in the paper are entirely our own. They do not necessarily reflect the views of the Embassy, Global Affairs Canada or any officers in these organizations.
References (22)
- et al.
Regionalizing results from a detailed CGE model: macro, industry and state effects in the U.S. of removing major tariffs and quotas
Pap. Reg. Sci.
(2007) - Baughman, L., Francois, J., 2009. Trade action – or inaction: the cost for American workers and companies, U.S. Chamber...
- Bridges Trade News Digest, 2017. French Elections: After Macron Win, Eyes Turn to Legislative Polls, Bridges - ICTSD...
- et al.
Dynamic general equilibrium modelling for forecasting and policy: a practical guide and documentation of MONASH
Contrib. Econ. Anal.
(2002) - et al.
Validating a detailed, dynamic CGE model of the U.S
Econ. Rec.
(2010) - et al.
Validation in CGE modeling
- et al.
Economy-wide effects of reducing illegal immigrants in U.S. employment
Contemp. Econ. Policy
(2011) - Dixon, P.B., Rimmer, M.T. and Waschik, R.G., 2017. Macro, industry and regional effects of eliminating Buy America(n)...
- et al.
Textile and apparel barriers and rules of origin: what's left to gain after the agreement on textiles and clothing?
J. Econ. Integr.
(2008) - et al.
Economy wide implications from US bioenergy expansion
Am. Econ. Rev.: Pap. Proc.
(2010)
Liberalizing NAFTA rules of origin: a dynamic CGE analysis
Rev. Int. Econ.
Cited by (8)
Toward green transition in the post Paris Agreement era: The case of Taiwan
2022, Energy PolicyCitation Excerpt :Flaig et al. (2016) evaluate the digital border controls with local data storage requirement by introducing the local content requirement mechanism as MCP in the CGE model. Dixon et al. (2018) use technical changes to simulate the effects of eliminating Buy America(n) requirements. The base year of our SAM is 2016.
Review on disaster indirect economic loss assessment
2023, Journal of Natural DisastersLocalization measures: A global perspective
2023, Local Content Requirements: Promises and PitfallsIt's Made in the USA, but it can't be branded 'Made in the USA'
2023, National Brands and Global Markets: An Historical PerspectiveBuying from Local Providers: The Role of Governance Preferences in Assessing Performance Information
2022, Public Administration Review