Supported by CIDER, the Collaboration for International Development Economics Research at Cornell University Supported by CIDER at Cornell
Labor & Migration · Vietnam

When Non-Compliance Gets Costlier, Do Firms Become More Formal?

By Alix Naugler October 7, 2026 5 min read

This is our first blog post for our Working Papers and Job Market Papers Series blog for 2026-2027.

Alix Naugler is a fourth-year PhD student in Applied Economics and Management at Cornell University whose research lies at the intersection of development, industrial organization, and agricultural economics.

Imagine a small manufacturing firm deciding whether to hire its tenth formally contracted, paid worker. The new hire could help the business grow. But in Vietnam, it could also trigger a regulatory requirement and a financial penalty if the firm fails to comply. Does the firm keep its formal workforce below ten, comply with the regulatory requirement, or find another way to respond?

In my working paper, The Cost of Non-Compliance: Firm Adjustments to Threshold-Based Labor Policies, I find that making non-compliance more costly can change firms’ workforce composition and which dimensions of formality they comply with. Rather than eliminating informality, greater regulatory pressure can reshape how firms operate within it.

A rule that depends on who counts as a worker

Article 119 of Vietnam’s 2012 Labor Code required firms with at least ten formally contracted, paid employees to prepare and register internal work regulations covering matters such as working hours, workplace safety, and disciplinary procedures. The ten-employee threshold requirement dated back to Vietnam’s 1994 Labor Code. What changed under the new legal framework was the financial penalty for non-compliance which increased five- to tenfold, from 1 million to 5-10 million Vietnamese dong (about 47 USD to 240-475 USD).

Crucially, only formally contracted, paid workers count toward the ten-employee threshold. In Figure 1, I refer to these as “counted workers” while workers without formal labor contracts and unpaid workers are “non-counted workers.” A firm can therefore employ more than ten workers in total while remaining below this threshold. As Figure 1 illustrates, two firms with the same total employment can face different regulatory obligations depending on the composition of their workforce. A firm can also be registered with the government without providing every worker a formal labor contract. In other words, being “formal” or “informal” is not necessarily an all-or-nothing distinction.

Figure 1: two hypothetical firms with 12 workers each. Firm A has 9 counted and 3 non-counted workers and does not reach the ten-employee threshold; Firm B has 10 counted and 2 non-counted workers and must register internal work regulations.
Figure 1. The figure illustrates two hypothetical firms with the same total employment but different workforce compositions. Formally contracted, paid workers count toward the ten-employee threshold while unpaid workers and workers without formal contracts do not count toward the threshold. Firms at or above the threshold must register internal work regulations or risk a financial penalty for non-compliance. Source: Author’s illustration based on Naugler (2026).

Firms could respond in several ways. Some might keep their formally contracted, paid workforce below ten by relying more on workers who do not count toward the threshold. Firms that cross the threshold face three options: submit the required internal work regulations and provide formal contracts to all workers, submit the internal work regulations while retaining workers without formal contracts, or fail to submit them and risk a financial penalty.

What firms did instead

To understand these responses, I use panel data from the UNU-WIDER Vietnam small- and medium-enterprise surveys, focusing on manufacturing firms near the ten-employee threshold before and after the policy change. By comparing firms initially just above and below the threshold in 2010 and 2014, I examine how their employment, formality, and performance changed after the financial penalty increased.

One might expect firms to avoid the regulatory requirement by stopping just short of employing ten formally contracted, paid workers. Yet I find no evidence of firms clustering just below the threshold after the financial penalty increased. Instead, firms appear to have responded in other ways.

First, firms changed their use of unpaid labor. After the policy change, firms initially just above the threshold experienced a relative decline of approximately 0.35 unpaid full-time workers compared with firms initially below it. This pattern is consistent with greater reliance on unpaid labor below the threshold—potentially including family members—rather than expanding the workforce that counts toward the threshold.

Second, firms shifted toward selective rather than comprehensive formalization. After the policy change, firms initially just above the threshold experienced a 15-18 percentage-point greater increase in the likelihood of being registered with the government without providing formal contracts to all their full-time workers, relative to firms initially below it. Thus, firms became more formal along this highly observable dimension of formality while informality persisted in their employment relationships. As Table 1 shows, a firm could therefore appear formal in government records without fully formalizing its workforce.

Table 1: firm formality by government registration and formal contracts for all full-time workers. Registered with contracts: fully formal; registered without contracts: partially formal (highlighted); not registered with contracts: partially formal; neither: fully informal.
Table 1. Firm formality depends on government registration and formal labor contract coverage among full-time workers. Fully formal firms meet both conditions, fully informal firms meet neither, and partially formal firms meet one or the other but not both. The highlighted cell represents government-registered firms that do not provide formal contracts to all full-time workers. Source: Adapted from Table 2 in Naugler (2026).

Third, these adjustments were accompanied by improvements in firm performance. After the policy change, firms initially just above the threshold experienced an estimated increase in gross profit per full-time worker of approximately 100-139 million Vietnamese dong (about 3,800-5,300 USD) relative to firms initially below it. I also find suggestive evidence of higher labor productivity for these same firms. These findings indicate that the effects of a higher financial penalty can extend beyond regulatory compliance, with firms experiencing performance gains even without comprehensive formalization.

Why government registration doesn’t tell the whole story

Governments often use firm registration rates to assess whether formalization policies are working. But registration captures only one dimension of formality and can overlook important changes within firms. Looking at government registration alone could therefore suggest greater progress toward formalization than workers’ actual employment arrangements reveal.

The design of the regulatory threshold also matters. Because only formally contracted, paid workers count toward it, firms can reorganize their workforce without necessarily reducing their total employment or fully formalizing. Policymakers therefore need to consider not only the cost of non-compliance, but how regulatory thresholds create incentives for firms to adjust in ways that may undermine broader formalization.

These findings concern Vietnamese manufacturing firms near the ten-employee threshold during a period when both the financial penalty for non-compliance and government inspections increased. Whether similar responses emerge in other settings remains an open question. Nevertheless, the results highlight an important consideration for policymakers: making non-compliance more costly for firms does not necessarily lead to comprehensive formalization. It may instead change where and how firm informality persists.

Keep reading

Leave a comment

Your email address will not be published. Required fields are marked *