In June 1992, BMW chose a peach orchard in Spartanburg, South Carolina, as the site of its first assembly plant outside Germany. The decision came after screening some 250 candidate locations worldwide. What followed was even more striking: 23 months from groundbreaking to the first car rolling off the line. BMW still cites it as the fastest such record in the auto industry.
Much of that speed had been locked in before the first shovel hit the ground. The state government bought up roughly 1,000 acres next to the airport from individual owners, assembled them into a single parcel and leased it for $1 a year. Incentive legislation was completed two months before the announcement. The contract even included a condition that the airport runway be extended so that Boeing 747 freighters could take off and land.
Sixteen years later, Volkswagen's site in Chattanooga, Tennessee, followed a similar pattern: 1,350 acres on the grounds of a former military ammunition plant, bought by the city and county, remediated, fitted with a highway interchange and certified as a TVA megasite. According to one account, when the due diligence team said from a helicopter that the trees made it hard to see the land, the mayor and the county executive redirected public works crews and finished clearing within 10 days. That was the backdrop to narrowing nearly 400 candidate sites to one in six months.
Both plants remain busy today. Spartanburg has become the BMW Group's largest plant and the top U.S. automotive exporter by value, and more than half of the BMWs sold in the United States last year came from there. Chattanooga has built the Atlas on the line that once made the Passat, and now builds electric vehicles. Even as the products changed, the two companies stayed on the sites they first chose for more than 30 years and 15 years, respectively.
The courtship that Korean companies are now receiving from state economic development agencies in the United States is also a story about a window that is open only for a while. Incentive budgets go to whoever commits first, power interconnection queues grow longer each quarter, and elections come around every two years.
Where does that speed come from? It may help to revisit notes from a U.S. business trip taken for a Korean company's new production base project. Managers from manufacturing, finance, human resources, logistics and real estate toured several candidate sites in one car, then sat down at a roundtable. "If the power doesn't work, none of it matters. Grid interconnection alone takes years." "Incentives there are half as much. Run it through NPV and the gap is large." "Did you check how many technical workers live within commuting distance?"
Everyone came with their own numbers, and everyone was right. In practice, evaluating a single site means looking at more than 40 indicators. In my experience, no site satisfies all of them. The site that ranks first on labor ranks near the bottom on power, and the place with the biggest incentives tends to have the slowest permitting. Meetings that wait for an answer satisfying everyone rarely end. That is why I tend to recommend settling three things before opening the candidate list.
First, decide what to give up. When looking for a home, you cut down on legwork by deciding in advance what to sacrifice among budget, school district and commute. Sites are the same. Agreeing before the site visits on whether the timing of power availability or labor costs comes first, and on which conditions disqualify a site no matter how good it otherwise looks, shortens the meetings. Writing down two or three disqualifying conditions alone cuts the long list in half. Reverse that order and each department pushes a different site, and the decision keeps getting pushed back.
Second, look at when you can break ground rather than at the price of the land. When does power arrive, how many months does permitting take, has remediation been completed, is there adjacent land left to expand into? Chattanooga had room to place a separate 440,000-square-foot supplier park at the plant gate, and seven companies making seats and axles feed the line directly from inside it. Spartanburg has carried out seven major expansions over some 30 years. A site chosen by calculating only your own plant footprint leaves no room for suppliers to move in.
Third, build enforcement mechanisms into the other side's promises. Incentive tables look impressive everywhere. The runway extension BMW asked for was actually delivered because it was a contract condition. For items that govern your own schedule, such as the power supply start date or the completion date of an access road, you can set out both a deadline and the consequences of nonperformance during negotiations. Everyone reads the clawback provisions attached to their own hiring and investment targets closely, but the dates are surprisingly often left blank on the counterparty's obligations.
What has to be secured in the United States is not just a single parcel. The right to build, the power to connect and local political approval come as one set. Assembling that set while the window is open requires something that has to be created before the long list of candidate sites. A single-page business case, with cross-departmental agreement on what to give up, may be what pulls forward six months in the United States.
She is...
A director at Cushman & Wakefield Korea and head of its Global Korea Desk, she focuses on overseas real estate strategy for Korean conglomerates and cross-border commercial real estate advisory. She previously worked at CBRE Korea and JLL Korea, and before that was a broadcast reporter on the economy desk at SBS CNBC, host of CNBC Asia's "Korea Report" and a correspondent at the World Economic Forum in Davos. She holds the Certified Commercial Investment Member (CCIM) designation and writes about the global flows of Korean capital through her newsletter "Korea Capital Decoded."
