Overview

In the 2014 Russia comparison, the clearest wider response was contraction among firms that continued operating, not statistically detectable excess exit. The targeted-export probability fell 8.55 percentage points relative to China; extensive-margin changes in non-targeted exports, other destinations, any exports, and any payroll were not statistically significant. Conditional on positive values, log non-targeted, other-destination, and total exports fell 0.2266, 0.2132, and 0.2816, while payroll fell 0.1117—described as roughly 20% lower total exports and 11% lower payroll. Shared production, recurring compliance, services, reputation, and uncertainty are plausible interfaces, not separately estimated causes.

Learning outcomes

The 2014 Russia restriction reduced targeted-export probability by about 8.55 percentage points relative to China, but not the probability of any exports or payroll differentially.
Conditional on positive values, total exports fell about 20% and payroll about 11%; at reported pre-period means, these correspond to roughly $75 million and $27 million per average treated firm.
Shared production, compliance, services, reputation, and uncertainty can transmit a shock, but the study does not separately estimate their contributions.

Keywords

Export controlsSpilloversPayrollProductionComplianceServices

How the Shock Travels Inside a Firm

The 2014 Russia evidence shows a distinctive kind of export-control response: targeted exports fell, but the clearest wider changes appeared among firms that continued operating. This chapter examines that pattern and the interfaces through which a destination-specific rule can affect non-targeted exports, services, and domestic payroll. The aim is not to choose one explanation prematurely, but to connect the measured outcomes to concrete production and compliance systems while keeping each proposed mechanism within its evidentiary limits.

The outcome is contraction, not disappearance

The strongest evidence of spillover comes from comparing the extensive and intensive margins. On the extensive margin, the Russia-targeted restriction reduced the probability that treated firms exported targeted products to the focal country by about 8.55 percentage points relative to the China comparison group. That is the direct response one would expect from a policy that made those transactions subject to licensing requirements with a presumption of denial.

The other extensive-margin estimates are different. The probability that treated firms exported non-targeted products to any destination fell by 2.12 percentage points, but the estimate was not statistically significant. The probability of exporting to destinations other than the focal country fell by 2.35 percentage points, also not statistically significant. The estimate for having any exports was a small, statistically insignificant decline of 0.94 percentage points. The indicator for any domestic payroll activity actually had a small positive coefficient of 0.43 percentage points, also statistically insignificant.

These results do not show that no firm changed its other products or destinations. A statistically insignificant average effect is not evidence that every firm was unaffected. It means the study does not find a sufficiently precise differential change in the probability outcomes for the treated group relative to the comparison group. The more visible response appears on the amount of activity among firms that remained positive exporters or employers.

On the intensive margin, the estimates are large and statistically significant. Conditional on nonzero values, the log value of non-targeted exports fell by 0.2266, and the log value of exports to destinations other than the focal country fell by 0.2132. The log value of all exports fell by 0.2816. The paper describes the total-export response as roughly a 20 percent decrease and, using the treated firms’ pre-treatment mean total export value of $376 million, translates it into an approximate $75 million decline per average treated firm relative to the pre-sanctions period.

Domestic payroll followed the same direction. Conditional on positive payroll, the log payroll coefficient was -0.1117, which the paper describes as approximately an 11 percent decline. At the treated firms’ pre-period mean payroll of $254 million, that is roughly $27 million per firm. The probability of having any payroll did not fall differentially, so the combined evidence is consistent with workforce reduction and operational downsizing among continuing firms rather than a wave of complete shutdowns.

This distinction answers a common but misleading question: “Did the firm exit?” A firm that stays open may still reduce production, cancel a product line, stop serving a destination, or cut technical and administrative staff. If a company’s fixed costs are spread across several products, a decline in export scale can also make the remaining lines less profitable. The absence of excess exit is therefore compatible with a meaningful loss of productive capacity.

The results also make a simple diversion story less convincing. If firms had merely stopped labeling a shipment as controlled or redirected the same goods through other destinations, one might expect activity in non-targeted products or other destinations to rise, or at least to remain stable as the targeted route fell. Instead, the study finds significant declines in the value of those broader categories among continuing firms.

Published by DeepMox

FAQ

Why did continuing firms matter in the Russia export-control results?

Because extensive-margin estimates did not show differential excess exit, while conditional intensive-margin values of exports and payroll fell significantly; staying open did not mean no economic contraction.

How much did total exports and payroll fall in the study?

Conditional on positive values, total exports fell by a log coefficient of -0.2816, described as roughly 20% or about $75 million per average treated firm, while payroll fell by -0.1117, roughly 11% or $27 million at the reported pre-period mean.

What mechanisms could transmit a targeted export control inside a firm?

Shared inputs and facilities, recurring compliance work, connected services and technical information, customer relationships, reputation, and uncertainty could transmit it; the study does not separately identify their contributions.

Previous10 minU.S. Export Controls: The 2014 Russia Test

Next13 minU.S. Export Controls: Weighing Costs and Security

U.S. Export Controls: How the Shock Travels

How the Shock Travels Inside a Firm

The outcome is contraction, not disappearance

FAQ