What a Targeted Export Control Really Costs

The subject of this series is the U.S. export-control system under the Export Administration Regulations and the domestic firms that operate within it. The final question is not whether a targeted restriction can stop the transaction it names; the evidence shows that it can. The harder question is whether the same rule also changes the wider firm, and how policymakers and firms should judge that trade-off without confusing measured effects, plausible mechanisms, and unmeasured security benefits.

The answer is a qualified yes. A control can be narrow in legal wording and broad in economic incidence because exports are not independent objects. Products share plants, inputs, engineers, and customer relationships. Compliance decisions share people, records, software, and management attention. Physical equipment may be bundled with software, engineering, repair, or data services. A destination-specific rule can therefore reach beyond the listed item without making every related item legally controlled.

That conclusion is not a claim that export controls are misguided. The purpose of the EAR includes national-security and foreign-policy objectives, and a restriction may be justified even when it imposes costs on U.S. firms. It is a claim about evaluation: stopping the named transaction is only one outcome. The policy should also be assessed for the wider activity it changes, the firms that bear that change, and the evidence available to distinguish necessary security costs from avoidable spillovers.

When the legal target meets the firm

The legal target is relational. Whether a product can be exported depends on more than what it looks like. The relevant facts can include technical specifications, classification, destination, end user, and end use. A product subject to the EAR but not listed on the Commerce Control List may be designated EAR99 and often may not require an explicit license, yet a restricted destination, recipient, or use can change the result. A listed item with an ECCN is not resolved by the code alone either; the destination and transaction conditions still matter.

That structure makes the firm a natural unit of analysis. The firm has to collect the facts, classify the item, screen the parties, decide whether a license or exception applies, retain the reasoning, and release or hold the shipment. A formal restriction on one route can generate work in sales, engineering, logistics, legal, compliance, information systems, and management. The work may protect the firm from an unauthorized export, but it can also delay a customer, reduce the value of a project, or make a related market less attractive.

The scale evidence shows why this is not a marginal issue confined to a handful of spectacular technologies. The share of U.S. export value subject to the EAR rose from about 33 percent in 2010 to about 60 percent in 2024. Much of the increase came from EAR99 shipments, whose share of all export value rose from about 23 percent to about 45 percent. At the same time, the paper reports that 95 percent or more of controlled export value did not require an explicit BIS license across the country groups in 2024, except for the different pattern in Group E.

Those figures fit together. The controlled perimeter can expand even when most controlled value does not end in a formal license application. Classification, screening, documentation, and monitoring still matter for shipments that are not licensed. The size of the controlled perimeter is therefore not the same as the intensity of every review, and neither is a direct measure of economic cost.

The distribution across destinations adds another correction to the usual picture. Group B countries accounted for about 77.3 percent of the value of EAR-subject exports in 2024, while Group D accounted for 18.9 percent and Group E for only 0.06 percent. Export-control exposure is not synonymous with comprehensive embargo exposure. Much of the trade occurs with destinations that receive more favorable treatment, but favorable treatment does not eliminate the need to establish the facts of a particular transaction.

The data are also a reminder that the measured perimeter is imperfect. Schedule B codes classify the physical characteristics of exported goods; ECCNs reflect technical characteristics, intended use, and reasons for control. One Schedule B code can be associated with several ECCNs, one ECCN with many Schedule B codes, and one physical code can appear on shipments with no listed ECCN. The mapping changes over time. A familiar trade category cannot be treated as a permanent legal answer.

This matters for the headline question because a mismeasured perimeter can produce a mismeasured cost. If analysts use physical product categories as a substitute for control classifications, they may count some ordinary shipments as controlled and miss others whose technical attributes matter. If they count license numbers as the whole compliance system, they will miss the decisions surrounding no-license-required shipments. If they infer the status of a foreign buyer from incomplete customs fields, they will mistake missing information for a cleared transaction.

What the Russia evidence establishes

The 2014 Russia episode provides the strongest causal evidence in the material. The policy, effective August 6, 2014, added eight ECCNs and 52 Schedule B codes to the exports to Russia subject to licensing requirements with a presumption of denial. The restriction focused on items that could support oil exploration or production in deep-water, Arctic offshore, or shale projects in Russia.