Macroblog

The US immigration system is a very important pipeline for young firms in the technology-intensive sectors to acquire talent as they grow. In particular, the H-1B allows US firms to hire foreign professionals in specialized occupations. But the odds for these young firms to obtain an H-1B visa for a qualified foreign worker are very low. In 2024, more than 750,000 applications competed for only 85,000 spots.

Some of the world's largest tech firms have resources to mitigate the impact of this randomness. They might have offices, subsidiaries, and legal departments around the world in which they can employ talent or choose to reapply for a visa the following year, as this research  shows. The early years of a startup, in contrast, are often defined by volatility, tight financial constraints, and intense competition for talent. For a small startup still testing its first product, a lost visa can increase the risk of missed deadlines, failed funding, and the end of a promising idea. (The Government Accountability Office provides additional evidence .) Losing one engineer could even derail an entire business plan.

Our research published in the Journal of Monetary Economics, "Skilled Immigration Frictions as a Barrier for Young Firms ," explores the intersection of this aspect of immigration policy and entrepreneurship. Consistent with the 2011 Government Accountability Office survey results, our findings suggest that US restrictions on skilled immigration are more of a burden to start-up firms and might contribute to an aging technology sector.

The puzzle of missing start-ups

During the past two decades, technological progress has accelerated at an extraordinary pace. Artificial intelligence and automation have redefined many industries. Yet by multiple measures, business dynamism in the technological sector—the churn of new firms entering and competing—has slowed. The share of high-tech firms among all existing new firms has declined since the early 2000s (see figure 1a). Important to our analysis, this decline in the share of high-tech startups cannot be explained by business consolidation driven by an increase in the market power of big-tech firms. In fact, there has been an increase in the share of small, less productive firms (with 1–19 employees) in the oldest age cohort of high-tech firms (ages 11 and older), which could arguably be the byproduct of a less competitive environment (see figure 1b). Notably, ample literature links declining business dynamism to slower economic growth (as this research  shows).

Coinciding with this decline in dynamism is a sharp tightening of skilled-immigration rules. In 2003, the US government issued 195,000 H-1B visas. Two years later, the cap dropped to 85,000, and it has remained there ever since, despite a significant increase in the number of applications, which may indicate an increasing demand for highly skilled foreign workers.

Winning the lottery increases the survival probability of young firms

To study whether immigration rules have contributed to the decline in start-up activity, we built a dataset linking three large sources of information: the US Citizenship and Immigration Services records on H-1B visa applications and lottery outcomes; Labor Condition Applications, which reveal how many foreign workers each firm sought to hire; and the National Establishment Time Series, a business-level registry tracking employment and firm survival across the country.

In our research, we followed more than 15,200 firms that were active between 2011 and 2020—large and small, young and mature, spreading across industry sectors but concentrated in technology and professional services. We use the randomness of the H-1B lottery as a natural experiment and compare the experience of those firms that "won" the visas they applied for with firms that, through purely random chance, did not.

Before discussing the model outcomes, we want to emphasize that all the results depend on the underlying assumptions of the modeling experiment, which our article explains in more detail. Different models with different assumptions might produce different results.

The results are striking. Winning the lottery significantly increases the likelihood of survival for young firms—those less than five years old—over the next five years. The average firm in a high-technology intensive sector that was granted all the visas it requested was roughly 3 percentage points more likely to remain in operation than one that lost all its applications. The effect of winning visas is strongest among start-ups with more than 10 employees—firms that have grown beyond the garage start-up stage but still lack the resources of a larger firm. In contrast, the lottery outcome has no measurable impact on average survival for firms older than five years.

These findings suggest that young firms—the ones most likely to develop new technologies and thus drive growth and innovation, as this research  demonstrates—bear the cost of H-1B visa caps, since access to skilled immigrants matters most to them.

Removing immigration barriers increases firm entry and productivity

We also built a quantitative model to assess how visa barriers might ripple through the broader economy. In the model, firms enter and exit markets, hire skilled and unskilled workers, and face three immigration hurdles: an up-front cost of navigating legal procedures, a per-worker visa fee, and a lottery of receiving the visa. Two key assumptions characterize native and foreign skilled workers in the model. First, they are imperfectly substitutable, as this research  shows. Second, following findings in the academic literature, foreign skilled workers are assumed to have higher relative productivity than domestic skilled workers. For example, a considerable literature (for example, see here , here , and here ) notes that foreign workers foster additional productivity gains. The fact that firms in practice do indeed hire foreign workers on H-1B visas and pay an additional cost to do so suggests that those workers might be more productive or offer some other benefit to these firms.

When these hurdles are removed in the model, firm entry increases by 65 percent in the aggregate, leading to fiercer competition in the market and increasing the average productivity of surviving firms by nearly half a percentage point—a substantial gain at the macroeconomic level. As young firms surge into the market, older and less productive firms exit, and the overall economy becomes more efficient. Our model sees a particularly large impact on the high-tech sector, with output expanding 26 percent.

The model, however, holds other economic implications. As skilled foreign labor becomes more available with the hurdles eliminated, competition in that segment of the labor market intensifies, lowering wages for native workers in high-skill occupations. However, as output rises as a result of higher productivity, prices fall and demand for complementary low-skilled labor grows, lifting salaries in low-wage occupations by roughly 12 percent.

We also evaluate some additional hypothetical changes to immigration policy in the context of our model. We find that productivity gains could be obtained if young firms have priority in the visa lottery, even if the aggregate H-1B cap remains the same. In this scenario, firm entry increases by 20 percent, and average productivity among surviving firms expands by nearly 1.5 percent (as older, less productive firms exit in large numbers). Importantly, the wage for native tech workers practically does not change as the H-1B cap remains the same. We also find that reduced administrative hurdles, such as the legal and filing costs that discourage small firms from applying, also produce measurable gains.

Takeaways

The construction of the US immigration system limits the growth of employment high-skilled foreign workers. Our study shows that more established firms perform better under the current immigration framework, while new firms that are more likely to be the next generation of innovators face substantial barriers to hiring foreign workers.

Our research suggests that the design of high-skilled immigration policy can have significant effects on aggregate productivity and economic growth. For example, even without expanding the total number of visas, reallocating them toward younger firms or lowering administrative costs could bolster start-up formation and productivity growth.

Note: The authors thank Lei Fang, Tom Heintjes, Paula Tkac, and Jon Willis for very helpful comments and Whitney Strifler for the interactive charts.

By Federico Mandelman, research economist and adviser in the Atlanta Fed's Research Department,

Mishita Mehra , assistant professor of economics at the University of Richmond, and

Hewei Shen , associate professor of economics at the University of Oklahoma