New York’s Amazon Debacle – Some Lessons

Predictably, Amazon’s decision to abandon its HQ2 project with 25,000+ future jobs in New York City  led to acrimony with opposing politicians and some commentators blaming the company for abandoning its effort.   Representative Alexandria Ocasio-Cortez, an early project opponent tried to re-frame her stance saying, “We do not have to settle for scraps in the greatest city in the world.  We deserve far more and can ask for more, and if they don’t want to negotiate, that’s their problem, not ours.”  But, on February 1st, she had argued,  “Nothing Amazon has said or done – including selling facial recognition technology to ICE & its intent to fight against worker unionization – would lead us to believe it could be a good or healthy neighbor for NYC,” 

A Cuomo Administration source to the Albany Times-Union blamed Amazon’s withdrawal on vocal opposition from Senator Michael Gianaris, whose district included the Long Island City site, and on Senate Majority Leader Andrea Stewart-Cousins. Stewart-Cousins’ decision to appoint Gianaris to the Public Authorities Control Board (PACB) “indicated to the company that the entire Democratic majority, and not just Gianaris was opposed to the project.”

Ocasio-Cortez’s and Stewart-Cousins’ messages to Amazon were hardly welcoming.  Their positions showed that New York could not promise that its governmental leaders would work together to make Amazon’s project a reality.  Amazon’s statement withdrawing from the New York project stated, “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…”

At best, the project’s opponents overplayed their hands in the negotiating process.  At worst, they showed a serious disregard for the creation of thousands of well-paying jobs for New York’s residents.  Perhaps they failed to recognize that Amazon and other businesses that could potentially locate within the state would be unwilling to risk timely project completion to time-consuming debates about incentive packages or about their fitness as corporate neighbors.

New York Had Business Location Advantages and Disadvantages

New York City had significant advantages as a business location for Amazon, perhaps the largest of which was the metropolitan area workforce, the largest in the nation, with its high concentration of information technology workers.  But, New York’s labor market advantages weren’t large enough to rule out other competitors.  Metropolitan areas like Washington, DC, Boston, Atlanta and Dallas had large enough labor pools to meet Amazon’s needs. In the wake of Amazon’s decision, other New York metros have offered themselves as alternatives but simply lack enough potential employees for a large information technology business headquarters.

But New York also had significant disadvantages.  Large projects in New York City are often difficult to develop because decision processes give vocal critics opportunities to delay or prevent them.  And, critics are often vocal in New York City.  In this case, the emergence of so-called “Progressive Democrats” who took office in the State Senate and Congress in the most recent election combined with neighborhood critics who feared the impact of the large Amazon project on their neighborhood.  The split among government leaders contrasted with northern Virginia, where according to Amy Liu of the Brookings Institution in the New York Times, “the Governor’s office, key state legislators and city and council officials worked together to address anticipated concerns from their constituency.”

Amazon’s proposed HQ2 project is unique – initially offering 50,000 jobs – more than any business expansion project in memory.  Economic logic tells us that Amazon had tremendous market power in making its decision.  Because of America’s Federal system, states can freely compete with each other to attract businesses.  Given that, it should not be surprising that there was intense competition between locations seeking to be home to the project.

States Use Incentives as Marketing Tools

States and localities market themselves by advertising the availability of tax incentives to businesses.  New York has been among the most aggressive in promoting them emphasizing the value of incentives available to companies that promise to create jobs.  State marketing marketing materials claim that “You’ll find all forms of tax incentives, business incentives and tax credits in New York State, all designed to benefit small or expanding businesses as well as film and TV production companies.” and that the state “offers new and expanding businesses the opportunity to operate tax-free for 10 years.”   In another place on the Empire State Development website, the text states, “Firms in the Excelsior Jobs Program may qualify for four fully refundable tax credits”   New York City has a set of  tax credits aimed at encouraging business investment in specific locations.

New York used incentives to attract businesses, spending billions of dollars to encourage technology development at Global Foundries in Saratoga County and at the Tesla Gigafactory 2 in Buffalo, among others. These technology related projects carried high risks. Tesla’s Buffalo venture, for example, is far short of the number of jobs that it promised to create and a film hub, developed in Syracuse, never attracted enough users to make it viable. A semiconductor manufacturer (ams AG) dropped plans to build a facility in Oneida County despite a substantial incentive package.

Amazon’s project was different, carrying relatively little risk for New York State.  Amazon is the market leader in internet commerce.  Though the deal was expensive – possibly more expensive than it needed to be, at least the tax credit incentives were performance based, and the capital grant would have contained recapture provisions if the company failed to meet job requirements.

Does Corporate Greed Make States Use Incentives?

Amazon received hundreds of proposals from localities hoping to secure the facility. New York’s winning $3 billion offer (about $1.7 billion from the state and $1.3 billion from the city) was not the largest. Offers from New Jersey ($7 billion) and Maryland ($8 billion) for the entire 50,000-job site were larger.

A number of commentators have blamed Amazon for being greedy for taking advantage of the incentives that states and localities create to lure businesses to operate within their borders.  For example, David Leonhardt in the New York Times wrote, “For years, companies have been getting the better of local governments — and taxpayers — by pitting them against one another. If a city or state won’t pony up cash or tax breaks, companies threaten to go elsewhere.”

Leonhardt’s  piece argues that states and localities are helpless victims of corporate greed.  In fact, governments are eager participants in corporate location competitions.  When New York got AMD (later Global Foundries) to locate a large semiconductor manufacturing facility in Saratoga County more than one billion dollars in total incentives were offered the company.   But, New York sought out semiconductor manufacturers, not the reverse.  The state had initiated a campaign to lure a facility to the state years before the deal was made.  The billion dollars in incentives was a lure, not the result of a bidding war.

Virtually all states and many localities market themselves to businesses.  Because location costs are often a factor in corporate decision-making, governments sell themselves as low cost locations, and use incentives, including tax incentives and cash grants, as marketing devices.

Incentives Waste Tax Dollars

Critics of these practices correctly point out that the use of business location incentives has several negative effects.  They create disadvantages for existing local businesses that do not receive them.  In cases where incentives go to companies that would have created jobs within the state without their benefit, they reduce revenues that would be available for needed government programs.

Advertising the value of tax incentives to businesses increases the cost to state residents of attracting new jobs.  New York advertises its Excelsior program as providing “A credit of 6.85% of wages per net new job [and]and an investment tax credit valued at 2% of qualified investments”  creating expectations by companies that like Amazon that they would be eligible for credits, even though the credits are discretionary and subject to a budgetary cap.

Given that the Excelsior Jobs Tax Credits that Amazon was offered were valued at $1.2 billion and that the company would have received $1.3 billion in tax credits from New York City, it is likely that they increased the size of New York’s offer beyond what it might have been if the credits had not been advertised by the state.

Amy Liu, writing in the New York Times reported that in contrast to New York’s approach, “Virginia offered Amazon $550 million in job-creation grants, which the company will receive only after delivering the proposed 25,000 jobs, with additional subsidies available if the company creates as many as 37,850 jobs….Virginia threw into the package more than $1 billion in additional taxpayer funds to build a pipeline of technical workers and improve transportation. This portion of the “subsidy” will not go directly into Amazon’s pockets but into Virginia schools, universities, and local agencies. It is nearly twice the amount offered to Amazon, a signal that investing in the local work force is more important than offering sweeteners to Amazon.”

Another New York practice that may have ballooned New York’s offer is the State’s practice of paying out incentives over long periods of time.  In Amazon’s case, the Excelsior Jobs Tax Credit would have been paid out over ten years.  10 years.  The state’s estimate shows that half the $1.2 billion in tax credits would be paid in years eight through ten of the benefit.  The  $505 million capital grant would have been paid over 15 years in proportion to the company’s annual investment.  There is evidence that paying incentives later rather than sooner wastes government funds, because companies heavily discount payments received far in the future compared to those to be received in the near term.  Timothy Bartik of the Upjohn Institute for Employment Research found that front loading incentives has more effects on firms’ location decisions per dollar of incentive costs because firms heavily discount the future.

Conclusion

The practice of providing financial incentives to businesses to locate within states and localities is destructive in a number of ways.  It invites states and localities to give away needed revenues in order to attract jobs.  It creates inequalities between companies that get benefits from government and those who don’t, creating the potential for crony capitalism.  It distracts government of focusing on improving infrastructure and investing in human capital.

Unfortunately, because of America’s federal system, states have the ability to set tax policy and use it to compete with each other for jobs.  Voluntary efforts by states to refrain from incentive competitions have been ineffective in the past because of the high stakes involved in foregoing potential job creation.

Because of its excessive reliance on tax incentives, New York likely pays more to companies for job creating projects than it could if it did not rely so heavily on them.  The incentives create company expectations that they can receive generous tax long term tax breaks if they invest in New York.  A policy that relied more on negotiated grants with strong claw-back provisions in the event of non-compliance with job creation requirements paid to reimburse corporate investments  would provide a better return on the state’s investment.

Although New York could have structured its Amazon incentive package more efficiently, the loss of 25,000 well-paying jobs (or more) to New York City is important both because of the foregone opportunity for area residents, and because it says to companies considering future projects in the city that New York’s political leadership is split between forces that welcome job creating projects and those who oppose them for a variety of reasons.  Perhaps the Governor and Mayor could have paved the way to a better reception from local political leaders by offering more involvement at an earlier stage, but given New York’s fractious political environment, it’s not certain that additional consultation would have eliminated objections from powerful local officials.

A version of this post appeared in the Rochester Beacon, titled “New York City’s Amazon Debacle”




How the State Senate Gerrymander Ultimately Hurt Upstate Residents

The November election brought a marked change in the composition of the New York State Senate that will have significant implications for upstate New York residents.  Tom Precious in the Buffalo News noted, “Majority party rules and minority party lawmakers are left with table scraps when it comes to funding and policy matters.  As a result, millions of upstate residents could see themselves lose the sole remaining seat at the table in closed door Capitol talks…”[1]

Kent Gardner wrote in the Rochester Beacon, “A New York State Legislature wholly controlled by downstate interests won’t stop sending education aid Upstate or close the state parks…. Complete control of the NYS Legislature by downstate reps will change things over time. Party aside, upstate has a lot at stake in this election.”

This post will seek to answer two questions:  Why did Republicans lose control of the State Senate, and why does upstate have so little majority party representation?

Republicans had controlled the State Senate from 1966, with only a brief break in 2012, when Democrats gained a majority and then splintered into two factions – one of which joined the Republicans to form a majority.  But in 2018, voters decisively gave control to Democrats, who will hold 40 of 63 seats.  The loss of Republican control will significantly reduce upstate’s representation in the party controlling the Senate, since only three seats outside the New York metropolitan area are held by Democrats.  In the most recent legislative session, 18 upstate Senators were in the Republican/Reform Democrat Senate Majority.

For many years, the Republican Senate majority was composed of a coalition of upstate, Long Island and Hudson Valley members.  In 2000, Republicans maintained control in the State Senate by winning 35 of 61 districts in all parts of the State – 17 in upstate New York, nine on Long Island, five in New York City, and four in the Hudson Valley portion of the metropolitan area.

In 2018, Senate Republicans won only three seats on Long Island, one in New York City and one in the Hudson Valley portion of the metropolitan area.  Only in upstate New York did the party increase its strength – to 18 seats.

The regions that moved towards the Democratic party – Long Island, New York City and the lower Hudson Valley — had substantial Democratic party representation by 2018.  Two-thirds of Senators from Long Island will be Democrats, all but one Senator in New York City and all but one in the Hudson Valley portion of the metropolitan area will be Democrats.

In upstate New York, only three of 21 State Senators will be Democrats.  Ironically, had the Senate’s legislative district apportionments more accurately represented voter party preferences, about half of upstate’s Senators would be Democrats.

Upstate’s majority party representation in the State Assembly will now be stronger than in the Senate.  Twenty-three of 48 upstate seats in the Assembly will be held by Democrats.[2]

Differences in Regional Party Affiliations

In New York City and in the lower Hudson Valley, Democrats and allied parties hold commanding registration advantages in 2018. Of those affiliated with a political party, 87% were Democrats in New York City and 62% in the lower Hudson Valley.  Long Island and Upstate New York are competitive – Democrats and allied parties are 52.5% of Long Island voters whore were affiliated with a political party in 2018, and in upstate New York, 52.7% were affiliated with Democrats or allied parties.

Upstate party registrations have shifted toward the Democratic party – from 47.3% to 52.7% between 2000 and 2018.  In 2000, in most upstate metropolitan areas Republicans and Conservatives were a majority of voters affiliated with a political party.[3]  By 2018, Republicans were in the majority in Binghamton and Utica, but Democrats (with Green Party, Working Family Party and the Women’s Equality Party) had majorities of party affiliated voters in the Albany-Schenectady-Troy, Buffalo-Niagara Falls, Rochester, and Syracuse metropolitan areas and in Tompkins County.

Within upstate New York, there are significant differences in the political affiliations of voters in metropolitan areas and outside them.  About one-third of upstate voters live outside the six metropolitan areas.  Almost 60% of these non-metropolitan residents who are affiliated with a party are Republicans.  But, residents of upstate’s larger metropolitan areas – Buffalo-Niagara Falls, Rochester, Albany-Schenectady-Troy and Syracuse lean Democratic.  Even the area’s smaller metropolitan areas – Utica-Rome and Binghamton show close to an even division between parties.

The significance of the split between registered Republicans and Democrats is somewhat weakened by the large number of upstate voters who are unaffiliated with a political party. 890,000 of the 3.84 million registered voters in upstate New York fall into this category.

Also note that the Independence Party was not included in this analysis because its candidate affiliations have been inconsistent.  Initially, In New York State, the party supported Tom Golisano, a wealthy Rochester businessman who ran on conservative a platform in 1994, 1998 and 2002.  More recently, the party has supported Democratic Gubernatorial candidates Eliot Spitzer and Andrew Cuomo.  The Independence Party is by far the largest splinter party in upstate New York, with 213,000 members.

Republican and Democrat Voting Strength in Upstate New York

This map, above (from Wikipedia), of the 2016 Presidential Election result suggests that Republican voters (in red) dominate upstate New York.  Other than metropolitan counties like Albany, Onondaga, Monroe and Erie, and a few counties in the Hudson Valley, most of upstate New York voted for Donald Trump, the Republican nominee.

This map looks like Republicans dominate upstate, but most of the red colored Republican Counties are sparsely populated, while the blue Democratic areas are population centers. And, the map is based on a single race. Voter preferences are more complex than can be captured in the results of a single face-off.

Contests for Congress and for Governor show that voting in elections in upstate New York is closely divided between Republicans and Democrats.  Republicans had more votes than Democrats upstate in the 2016 election for President and the 2018 election for Governor, but more votes were cast for Democrats in the 2012 Presidential election and the 2018 Senate and House of Representatives elections.

Why Upstate State Senate Representation Does Not Reflect Voter Preferences

Upstate representation in the State Senate does not reflect underlying party preferences as measured by party registration or voting behavior.  Democrats only hold three of twenty-one upstate State Senate seats, even though races at the State and Congressional levels have been highly competitive, with Democrats winning 5 of 9 upstate congressional districts in 2018, and 23 of 48 Assembly seats as well as gaining more votes in one of two presidential elections and in the most recent U. S. Senate election.  Why is the Senate unrepresentative of upstate party preferences and election performance?

When I began researching this question, I assumed that the primary cause of the difference in electoral results in upstate State Senate districts and other elections had to be legislative district gerrymandering.  After all, what else could explain the fact that Republicans control 18 of 21 districts, even though party preferences are about evenly split between the parties?  Though there is clear evidence that the controlling parties in the Legislature structured districts to favor members of their own parties, another factor appears to be important.  Republicans continue to succeed in electing candidates in upstate Senate districts where there are more Democrats than Republicans.

Gerrymandering

The State Legislature has historically controlled the redistricting process, creating gerrymandered districts that favored the parties in control.  In 2010, Republicans controlled the State Senate, and Democrats controlled the Assembly.  The result was a series of gerrymandered districts.

For example, State Senate District 50 in the Syracuse Metropolitan Area was carefully constructed to include suburban Republican votes, while excluding Democratic votes in the city of Syracuse.  The seat was held for many years by Senator John DeFrancisco until his retirement this year.  In the 2018 election, another Republican, Bob Antonacci, was the victor.

Assembly District 101, which looks like a worm that connects the town of Montgomery, just west of Newburgh with New Hartford – a suburb of Utica – is another example. The district, which carefully avoids nearby cities (with higher Democratic enrollment), is about 150 miles long and includes parts of Oneida, Herkimer, Otsego, Delaware, Ulster and Orange Counties! It is held by Republican Brian Miller.

Gerrymandering in the legislative apportionment process attempts to make one party waste as many votes as possible by concentrating voters in that party in overwhelmingly one-party districts, while spreading votes in the party benefiting from the gerrymander across districts in smaller, but safe majorities.

Since the mere requirement of population equality does not necessarily result in the creation of legislative districts that come close to reflecting the political preferences of residents, political parties within legislatures have taken advantage of the opportunity to maintain control, often creating legislative majorities that are much stronger than voter preferences would indicate, or in some cases, maintaining control despite the fact that only a minority of voters prefer them. Misalignments between voter preferences and public policy result from partisan gerrymanders.

Additionally, upstate Senate districts on average have fewer residents than downstate districts in order to strengthen upstate’s presence.  The difference in district sizes increased the region’s representation from 20 to 21 seats.  Under the law, legislative district populations may vary by 10%.

In upstate New York, State Senate districts were engineered to break up metropolitan areas into a series of districts that also contained large numbers of non-metropolitan residents, predominantly Republican voters.  The result has been to weaken the ability of democratic leaning metropolitan area voters to elect representatives who shared their political preferences.

For example, the Albany-Schenectady-Troy metropolitan area has 887,000 residents.  With 63 Senators, each Senate District should have approximately 315,000 residents.  The area has enough residents for almost three State Senate districts.  But the most recent Senate Reapportionment in 2016 split the metropolitan area into pieces of four State Senate districts that encompass large rural areas outside the metropolitan area.  In the process, three of the area’s primary cities – Troy, Schenectady and Saratoga Springs were each split into two Senate districts, disregarding their boundaries.

Similarly, the Rochester MSA, with a population of 1.1 million has enough residents to populate three and one third Senate districts but has been split into five Senate Districts including substantial non-metropolitan populations.

An analysis by Jeremy Creelan and Allison Douglas of the Rockefeller Institute, “New Tools to Challenge Partisan Redistricting in New York State”[4]  measured the extent of partisan gerrymandering in New York and examined the reviews of other authors and came to the conclusion that the extent of partisan gerrymandering in New York favoring Republicans was substantial, quoting an analysis by Simon Jackman of Stanford University that concluded that New York’s districts were “the most Republican favoring out of any state…”[5]

Democratic Underperformance

Although State Senate districts were structured to disadvantage Democratic party candidates by mixing Democratic majorities in metropolitan areas with rural, Republican voters, Democrats have not elected as many State Senators as would be expected from upstate Senate districts’ partisan makeups.  Eight of 21 upstate Senate districts have more Democratic voters than Republicans, but only three districts are represented by Democratic State Senators.  Several factors may account for Democrats’ underperformance.

  • The natural advantage conferred by incumbency is reinforced by gerrymandering. In 2018, 18 of 21 Senate contests involved Republican incumbents. The Democrats didn’t even field a candidate in 7 of these races. The three winning Democrats were incumbents. (A similar dynamic plays out in the Assembly. Eighteen of 48 Assembly contests were uncontested.)
  • There is a large fall-off in voting between statewide candidates, and those for the State Senate – 15% of those who voted in the election for U. S. Senate did not vote for a State Senate candidate. This may reflect Democratic voters’ decision to abstain from voting when the race is uncontested or manifestly uncompetitive.

Implications

Viewed purely with “sectional” eyes, upstate New York’s influence in Albany may have been greater because Republican gerrymandering granted upstate control of one of the Legislature’s two houses.  But that power meant that the views of about half of upstate’s residents – Democratic voters who make up the majority of residents of metropolitan areas – were underrepresented in the State Legislature.  The upstate Republican leadership of the State Senate primarily represented the interests of conservative rural and suburban residents.  Upstate city residents and people whose policy views corresponded with Democratic party positions received little Senate attention.

With Democratic party control of the State Senate, upstate now has less majority party representation than it would if Senate districts had not been gerrymandered in favor of Republicans, weakening the region’s voice in the State Capitol.

To be sure, fairer legislative districts would not ensure particular outcomes in State Senate races, because voters do not vote for parties in legislative elections any more than they do in elections for Governor or the U. S. Senate.  The large difference in the percentage of upstate voters who favored Andrew Cuomo (46%) against Republican Marcus Molinaro compared with the support given Kristin Gillibrand (57%) against Republican Chele Farley illustrates this point.  In fact, in much of upstate New York, party affiliations are relatively evenly divided, with electoral results in Statewide elections favoring each party in different races.

Reapportionment after the 2020 Census

Reapportionment after the 2020 Census is likely to create legislative districts that more accurately represent the partisan preferences of New York residents than past apportionments.  In 2012, New York voters passed a Constitutional amendment that reforms the apportionment process and institutes new Constitutional requirements for district representativeness.  The process is a mixed bag, containing features that could work against accurate voter representation, and others that could work for it.

The amendment sets up a ten-member redistricting commission with four members appointed by each of the major parties’ legislative leaders plus two additional members selected by the eight.  The amendment also requires that any plan developed receive the support from members nominated by both political parties.  The State Legislature must vote upon the plan submitted by the Commission without amendment. If two successive plans from the Commission fail to receive Legislative and gubernatorial approval, the Legislature is then empowered to present its own plan. As legislative approval is required, there will be an unnamed “third party” involved in the redistricting effort, the “Party of the Incumbency.” Logrolling and trading to preserve the prerogatives of incumbents and secure the required approvals may work against representation of voter preferences within legislative districts.[6]

The Constitutional amendment includes a prohibition against partisan gerrymandering, “[d]istricts shall not be drawn to discourage competition or for the purpose of favoring or disfavoring incumbents or other particular candidates or political parties.  This important provision is intended to prevent the redistricting commission from manipulating district boundaries to advantage particular parties and candidates and  creates a constitutional basis for a court challenge to a legislative districting plan that is gerrymandered.

Because of the statewide shift in voter preferences to the Democratic party, Republicans are unlikely to retake control of the State Senate, at least in the near future.  Upstate New York is unlikely to have as much influence as it had in the past in the State legislature.  But if the  passage of the 2012 constitutional amendment achieves the goal of representational equity,   the Senators who represent it will be more likely to represent the policy preferences of upstate residents.

(Note:  An abridged version of this post appears in the Rochester Beacon, titled, “The Cost of Gerrymandering”)

[1]Potential State Senate Power Shift Offers Big Implications for Upstate,” Tom Precious, Buffalo News, October 7, 2018.

[2] Upstate districts outside the New York metropolitan area are districts 101-103 and 106-150.

[3] The Independence Party is excluded from the analysis.

[4] Jeremy Creelan and Allison Douglas, New Tools to Challenge Partisan Redistricting in New York State?”, Rockefeller Institute of Government, August 31, 2017, https://rockinst.org/issue-area/new-tools-challenge-partisan-redistricting-new-york-state/

[5] Simon Jackman, “Assessing the Current Wisconsin State Legislative Districting Plan,”  July 7, 2015, http://www.campaignlegalcenter.org/sites/default/files/Jackman-WHITFORD%20V.%20NICHOL-Report_0.pdf

[6] Creelan and Douglas, op. cit.