Not what I would have done, but the fact that citizen opposition led to the Amazon HQ2 NYC venture being dropped is important to acknowledge

From the BBC News article "Amazon cancels New York City campus plan":

Amazon has said it will not build a new headquarters in New York, citing fierce opposition from state and local politicians. The dramatic turnabout comes just months after the firm named New York City one of two sites selected for major expansion over the next decades.

City and state leaders had agreed to provide about $3bn (£2.3bn) in incentives to secure that investment. Those subsidies had prompted fierce backlash in some quarters.

Amazon said its plans to build a new headquarters required "positive, collaborative relationships with state and local elected officials who will be supportive over the long term".

It said: "A number of state and local politicians have made it clear that they oppose our presence and will not work with us to build the type of relationships that are required to go forward with the project we and many others envisioned.

"We are disappointed to have reached this conclusion."

======
While I consider myself a progressive leftist, the reality is that my politics and recommendations are mediated by the reality of working within the capitalist system and aiming to achieve as much social and public good as possible, given how the private sector does most of the actual doing when it comes to revitalization-focused real estate development.

So yes, I understand and support, depending on the case, tax and other incentives to facilitate certain kinds of development.

Offer the best package of tax incentives for the locations and projects that do the most to spur change and additional private investment. Instead of offering tax incentives on a first-come, first-served basis, without much in the way of criteria or focus, the best set of incentives should be targeted to those locations--anchor projects-- that are determined to have the most potential for spurring additional private investment because of they example of "quality" that they provide, which without the incentives, would not occur for many years, given the current investment climate and perceptions of the submarket.

Only give tax abatements to quality projects. Similarly, tax incentives should be directed to projects that are special rather than to projects that are ordinary and/or sub-optimal.

That means don't do what they do in Tennessee, where reports say that on average incentives are about $1.2 million per job ("State's largest tax credit for biz costs $1.2M per job, study finds," Nashville Tennessean) and Illinois, where an analysis of the State's main tax credit programs by the Chicago Tribune found that most of the projects did not generate significant economic return ("Chicago Tribune investigation on state tax incentive deals with corporations").

Amazon.  On that basis, from a benefit-cost analysis standpoint, providing financial incentives to Amazon, even $1 billion or more, is beneficial because of their obvious track record in meeting the contract terms in terms of actual jobs and buildings developed.

From "An example when I may disagree with Richard Florida: incentives for landing the Amazon HQ2":

... wrt this specific project I will say:

1. The cost/benefit is a helluva lot better compared to a stadium or other sports facility
2. It doesn't involve competition for the relocation of an already existing facility (e.g., Mercedes Benz moving to Atlanta from New Jersey; GE moving from Connecticut to Boston; etc.)
3. The firm seeking the incentive package has a high quality track record making the proposal significantly less risky compared to a business not already operating in the US or a start up firm or a project seeking to redevelop a brownfield site.
4. There are clear, monetizable benefits from landing this project.

Amazon is a much different case than the typical one.  Not just Foxconn in Wisconsin, as the firm has a history of not following through on "commitments" but more generally.  And definitely not for sports stadiums and arenas.

And the business, depending on what it is (warehouse jobs no, business development and technology jobs yes) has the opportunity to be leveraged to further develop related education disciplines and business activity, like how Virginia Tech is developing a computer sciences related school adjacent to the future HQ2 project in Crystal City/Arlington County, Virginia ("A look at Virginia Tech's planned $1 billion Potomac Yard campus").

Why not locate in an area that needs an economic boost?  Me, for social and other reasons, I would have suggested Baltimore -- which didn't make the cut; or Newark, which did, or Philadelphia.
But given that Amazon's HQ2 search hasn't focused on criteria I would have thought that mattered:

-- "The Amazon second headquarters "****show": Part 1 | Where could it go?," 2017
-- "Amazon second headquarters list of finalists," 2018

that wasn't the way the decision went.

Amazon's choice of Long Island City for half of the second HQ.  Amazon originally chose to split the new HQ2 between Crystal City in Arlington Virginia, across the Potomac River from DC, and a commercial district in Queens, New York City.

That made sense to me, because the DC area has large government customers as well as agglomeration economies when it comes to Internet backbone and related services, and because New York City is a center for both financial services and advertising, clusters where Amazon aims to generate more business.

The building that Amazon intended to locate its LIC-based HQ2 activities.  

Both places seemed to be selected on the basis that they had a lot of commercial space immediately available.  But yes, I was surprised that Amazon chose Long Island City in Queens, New York City.

Pushing higher housing costs up even higher.  Not because LIC is a bad place to do business, but because one of Amazon's originally stated reasons for not continuing to expand in Seattle was because of the negative effects on housing and the cost of doing business.

New York City is the highest cost place to do business in North America.  Sure, Queens is cheaper than Manhattan, and the former Citicorp building in Long Island City has plenty of empty space, virtually all that Amazon needs.  But there will be massive effects on the cost of housing and other negative effects.

Opposition.  There has been lots of opposition to the deal for New York City, both against incentives to be provided by the State of New York and separately by the City of New York.  So Amazon has backed out.

While I would have argued in favor of the deal, were I living in New York City, I can see why other people disagree and would fight the deal.   While I don't agree with them, I am amazed that this opposition caused Amazon to change its mind.

Interestingly, more than 50% of the population in a recent poll supported the incentive package ("Majority of New Yorkers support Amazon HQ2 deal, according to poll," CNY Central).

Chance to make a better decision.  But then, because of the impact on the housing market specifically, I thought that Amazon choice of New York City was wrong-headed.

And maybe now there is an opportunity to direct the investment to a location where the effects can be mediated and have extra-normally positive effects, by contributing to business development in a place that is underdeveloped, rather than a place that is already economically overdeveloped by comparison.