Politicians across the U.S. are scrambling to entice Amazon to build its second headquarters — and the 50,000 jobs that come with it — in their region. In fact, the world's largest online retailer announced this week that it has received 238 proposals for Amazon HQ2.

In this frenzied and competitive environment, studies that analyze government incentives to attract companies loom large. The big question: Are these economic development subsidy packages worth it?

A Good Jobs First study last year found that local and state governments have paid an average $658,427 per job in 386 "megadeals" of economic development packages (valued at $50 million and up) to companies since 1976, and received less than that amount in income taxes, sales taxes, property taxes or other revenue. In June 2013, the group found 264 megadeals with a $456,000 per job price tag.

"We knew that megadeals were getting more expensive," said Greg LeRoy, head of Good Jobs First, the national nonprofit based in Washington, D.C.

What accounts for this rising price? The more recent megadeals are "extremely capital intensive," LeRoy said. In sum, there is a high ratio of expenditure on capital over labor due, e.g. machines to people, from highly automated company operations.

Taxpayer financial assistance packages from local and state governments studied range from lower hotel, property and sales taxes to cheap and even free land. The top 10 recipients of tax dollars read like a who's who of corporate America. Boeing tops the list, followed by Alcoa, Foxconn, General Motors, Ford Motor, Sempra Energy, Nike and Intel.

The main portion of the Washington state legislature's $8.7 billion package to Boeing, an aircraft manufacturer, was a 16-year extension of the tax breaks that it received in 2003, with the incentives available to the company's suppliers, too. The Wisconsin legislature passed and Gov. Scott Walker inked a $3 billion subsidy package to attract Foxconn’s flat-screen factory to the Badger State in 2017.

"Spending so much on so few companies, a strategy often called 'buffalo hunting' in economic development circles, is both risky and wasteful: deals with such high per-job price tags can never break even from a fiscal perspective, much less generate a positive taxpayer return on investment," according to the study, which LeRoy co-wrote with Thomas Cafcas. "That is, workers at such facilities will never pay that much more in taxes than public services they and their dependents consume."

What about economic development subsidy packages and smaller companies? Small businesses created 48,000 jobs in August and 50,000 in July, according to Automated Data Processing Inc.

"Incentives that focus more on small- and medium-sized businesses, and that provide upfront services such as customized job training and manufacturing extension services are likely to be more effective," said Timothy J. Bartik, senior economist at the W.E. Upjohn Institute for Employment Research in Kalamazoo, Mich.

The Good Jobs First report arrives against the backdrop of the Oct. 19 deadline for cities and states to submit bids for Amazon to build its $5 billion headquarters. Governors and mayors across the U.S. have dangled financial incentives before the Seattle-based internet giant to attract its second headquarters.

Amazon estimates that "its investments in Seattle from 2010 through 2016 resulted in an additional $38 billion to the city’s economy — every dollar invested by Amazon in Seattle generated an additional $1.40 for the city's economy overall." According to Bartik, however, the return on taxpayer incentives has important nuances.

"Government is not a business trying to make money for taxpayers or itself," Bartik said. "Rather, the government is trying to spend money to provide public benefits."

Government spending for public infrastructure such as roads and streets is an example. Such expenditures have related employment impacts.

"In the case of economic development incentives, the main public benefit is labor market benefits," Bartik said. "If these incentives tip location decisions, they will raise employment to population ratios in the local labor market and increase real wages."