Openness, Trade, and Knowledge Diffusion

Why Openness Can Affect Productivity

A country may gain access to technology developed elsewhere without having produced that technology itself. Imported machinery can embody a new production method; a foreign-owned firm may bring management practices and technical standards; and a domestic supplier may learn by serving international customers. These are possible routes through which openness can transmit knowledge and raise productivity.

Openness does not guarantee a productivity gain. Firms need the skills, infrastructure, finance, and organizational capacity to adapt what they encounter. Some knowledge is easy to copy from a product or manual; other knowledge is tacit, acquired through training and repeated production. Gains can also vary across firms, industries, regions, and workers. The central task is therefore to distinguish a channel that can transmit knowledge from the conditions that determine whether it is adopted and who benefits.

Definition: Knowledge diffusion is the movement or transfer of useful ideas, techniques, designs, or production practices among people, firms, or countries. Openness refers here to economic connections across national borders, including trade in goods and services, cross-border investment, and international research or production networks. Openness can facilitate diffusion, but it is not itself a measure of successful adoption.

Channels of Cross-Border Diffusion

Trade can transmit knowledge through several routes. Imported capital goods may contain technology that domestic producers could not readily develop on their own. Imported intermediate inputs can provide better quality, new features, or more reliable production methods. Exporting can expose domestic firms to demanding buyers, international standards, and production requirements that encourage learning. These effects depend on whether firms can understand and use the new information or inputs.

Cross-border investment can provide another route. A multinational firm may bring equipment, management systems, worker training, and quality-control methods to a host country. Domestic suppliers may learn by meeting the multinational’s standards, while workers may carry skills to other employers. These are potential spillovers: benefits that reach parties beyond the investing firm and are not necessarily fully captured by it. They are not automatic. A foreign firm may operate with few domestic links, or local firms may lack the resources to imitate its methods.

Research partnerships, professional networks, and the movement of skilled workers can transmit knowledge directly. International collaboration may connect researchers to ideas and equipment unavailable locally. The practical effect still depends on local institutions and capacity. A useful distinction is between access to a technology and absorptive capacity, the ability to recognize, adapt, and apply knowledge from outside.

Key distinction: A larger foreign technology frontier creates an opportunity to catch up, not a guarantee of catch-up. The domestic productivity response depends on the strength of the transmission channel and the ability to absorb what is transmitted.

A Simple Model of Technology Diffusion

A compact model helps organize this distinction. Let \(A_t\) be domestic productivity at the start of period \(t\), and let \(A^*\) be a fixed foreign productivity frontier, with \(A_t \leq A^*\). Let \(m\) measure effective exposure to foreign knowledge during the period, where \(0 \leq m \leq 1\), and let \(\theta\) measure domestic absorptive capacity, also between zero and one. In this stylized model, the part of the productivity gap closed in one period is the product \(\theta m\).

Assume that the frontier is fixed during the period, exposure and absorptive capacity are constant within it, and no other source of domestic productivity change is included. Then the model is:

A_{t+1}=A_t+\theta m(A^*-A_t)

This is an illustrative convergence equation, not a universal empirical law. It isolates a diffusion mechanism: if either exposure or absorptive capacity is zero, this channel produces no gain. If both are positive, domestic productivity moves toward—but does not pass—the fixed frontier. Other forces, such as domestic innovation or a changing frontier, would need to be modeled separately.

The equation also gives a useful result about the remaining gap. Define \(D_t=A^*-A_t\). Subtracting the next-period productivity from the fixed frontier gives:

\begin{aligned} D_{t+1} &=A^*-A_{t+1}\\ &=A^*-\left[A_t+\theta m(A^*-A_t)\right]\\ &=(1-\theta m)(A^*-A_t)\\ &=(1-\theta m)D_t \end{aligned}

Since \(0 \leq \theta m \leq 1\), the next-period gap is between zero and the current gap. The model therefore says that a greater effective diffusion rate closes a larger share of the gap in that period. It does not say that every firm adopts the technology or that the resulting gains are shared equally.

Worked Example: Calculating a One-Period Diffusion Gain

A hypothetical domestic industry has a productivity index of 90, while a relevant foreign frontier is 150. Trade links expose the industry to the foreign methods with \(m=0.25\), and domestic absorptive capacity is \(\theta=0.40\). Under the simple diffusion model, calculate next period’s productivity and the share of the initial gap that remains.

Method and assumptions. Treat the frontier, exposure, and absorptive capacity as fixed for this one period, and exclude other productivity changes. The initial gap is:

$$ D_0=A^*-A_0=150-90=60 $$

The fraction of that gap closed is:

\theta m=0.40(0.25)=0.10

The productivity increment is \(0.10(60)=6\), so next period’s productivity is:

$$ A_1=90+6=96 $$

The remaining gap is \(150-96=54\), which is \(54/60=0.90\), or 90% of the initial gap. The method avoids treating the exposure measure \(m\) alone as the fraction of the gap closed; absorptive capacity also matters. Check: the next-period gap from the model is \((1-0.10)(60)=54\), matching the direct calculation.

Level Gains and Ongoing Growth

A technology transfer can raise productivity once, while continued growth describes repeated increases over time. These are different effects. If a one-time adoption raises a productivity index by 5%, and productivity then grows by 3% per year for two years, the level change is multiplicative: first apply the adoption factor, then apply each year’s growth factor. The annual growth rate is not added to the one-time percentage as if both were a single-period change.

Worked Example: Combining Adoption and Annual Growth

A hypothetical manufacturer has a productivity index of 200. Adoption of a foreign production method raises its index by 5% at the start. Productivity then grows by 3% annually for two years. Calculate the final index and compare it with the index under the same annual growth but without the adoption. Round only the final indices to two decimal places.

Method and assumptions. Treat the adoption as a one-time level increase and the 3% rate as an effective annual productivity growth rate applied at the end of each year. The adoption factor is \(1.05\), and the two annual growth factors are each \(1.03\). Thus:

$$ 200(1.05)(1.03)^2 =200(1.05)(1.0609) =222.789 $$

The final index is 222.79. Without adoption, the index would be:

$$ 200(1.03)^2=200(1.0609)=212.18 $$

The difference is \(222.789-212.18=10.609\), or 10.61 index points to two decimals. Check: the adoption adds \(200(0.05)=10\) points immediately; those 10 points also receive two years of growth, giving \(10(1.03)^2=10.609\). The common mistake is to add rates or use an incorrect square of 1.03. Compounding gives \(1.03^2=1.0609\), so the final index is 222.79, not 222.66.

Why Aggregate Gains Can Be Uneven

An increase in national or industry productivity is an aggregate statement. It does not imply that every firm adopts a technology, that all sectors expand, or that workers receive equal gains. Firms that already have trained staff, suitable equipment, and access to finance may use imported technology quickly. Smaller or less connected firms may not. Within a firm, workers whose tasks complement new technology may benefit differently from workers whose tasks are replaced or reorganized.

Trade and investment can also change which activities expand. A sector that gains access to improved inputs may become more productive, while another sector faces stronger competition and contracts. Workers and regions differ in how easily they can move into expanding activities. In the short run, adjustment costs—such as retraining, relocation, or unused equipment—can make the distribution of gains especially uneven. Aggregate productivity can rise even when particular groups experience disruption or little improvement.

A calculation of productivity does not by itself establish who receives higher wages, profits, or consumer benefits. That further conclusion requires assumptions about competition, bargaining, labor mobility, ownership, and how productivity changes affect prices and employment. Keep the measured outcome clear: productivity is output per unit of input, while the distribution of income is a separate question.

Worked Example: Aggregating Unequal Sector Gains

A hypothetical economy has two sectors and fixed employment over the period. Sector A employs 60 workers, each producing 100 units; sector B employs 40 workers, each producing 80 units. New cross-border production links are associated in this scenario with productivity increases of 10% in A and 5% in B. Calculate the change in total output and the percentage change in output per worker.

Method and assumptions. Hold employment and hours fixed, and apply each sector’s stated productivity change to its output per worker. Initially, sector A produces \(60(100)=6{,}000\) units and sector B produces \(40(80)=3{,}200\) units. Initial total output is \(6{,}000+3{,}200=9{,}200\) units for 100 workers.

After the changes, A’s output per worker is \(100(1.10)=110\), and B’s is \(80(1.05)=84\). Their outputs are \(60(110)=6{,}600\) and \(40(84)=3{,}360\), respectively. Total output becomes \(6{,}600+3{,}360=9{,}960\) units, an increase of \(9{,}960-9{,}200=760\) units.

Initial output per worker is \(9{,}200/100=92\); final output per worker is \(9{,}960/100=99.6\). The percentage increase is:

\frac{99.6-92}{92}\times100 =\frac{7.6}{92}\times100 \approx 8.26\%

Sector A’s productivity rises by 10%, while sector B’s rises by 5%, even though aggregate output per worker rises by about 8.26%. The method weights the changes through actual sector output and employment, rather than taking an unweighted average of 10% and 5%. Check: the aggregate increase is \(760/9{,}200\approx 0.0826\), the same percentage. This calculation says nothing by itself about the distribution of wages or profits.

Interpreting Openness Carefully

When evaluating a claim that openness raised productivity, separate the proposed channel from the observed outcome. Ask what knowledge or technology could have crossed the border, which firms or workers were exposed, and what capabilities made adoption possible. Then distinguish a one-time level improvement from a sustained growth-rate change. A technology can produce a level gain without permanently increasing the rate of innovation.

It is also important not to confuse association with a proven causal effect. Productive firms may be more likely to export or attract foreign investment in the first place. A simple before-and-after comparison may therefore mix the effect of openness with pre-existing differences or other changes occurring at the same time. The diffusion model above clarifies a mechanism and its assumptions; it does not, on its own, identify an empirical causal effect.

Key takeaway: Trade and cross-border investment can carry knowledge through imported inputs, exporting, multinational production, supplier links, and research networks. Productivity gains depend on exposure and absorptive capacity, and aggregate gains do not establish that benefits are equal or that every group gains.

This treatment is aligned with the SOA VEE Economics (Validation by Educational Experience) syllabus: focus on the mechanisms, the conditions under which they operate, and the distinction between aggregate productivity and the distribution of gains.

Check Your Understanding

Use the stated assumptions in each problem, and check that each calculation matches the economic quantity requested.

A domestic producer has productivity index 90 and faces a fixed foreign frontier of 150. If exposure is \(m=0.25\) and absorptive capacity is \(\theta=0.40\), use the one-period diffusion model to find next period’s index. (Answer: 96)
A fixed foreign frontier is 140. Domestic productivity is 100, exposure is \(m=0.20\), and absorptive capacity is \(\theta=0.50\). How many index points does diffusion add in one period, and what percentage of the initial gap remains? (Answer: 4 index points; 90% of the initial gap remains)
A firm’s productivity index is 250. A one-time technology adoption raises it by 4%, followed by annual productivity growth of 2% for three years. What is the final index, rounded to two decimal places? (Answer: \(250(1.04)(1.02)^3=275.91\))
Sector A has 80 workers producing 120 units each; sector B has 20 workers producing 60 units each. With employment fixed, A’s productivity rises by 5% and B’s by 15%. What is the percentage change in total output per worker, rounded to two decimals? (Answer: initial output \(=10{,}800\), final output \(=11{,}460\); increase \(=660/10{,}800\approx 6.11\%\))
A domestic industry starts at productivity 70, with a fixed frontier of 130. Exposure is \(m=0.20\) and absorptive capacity is \(\theta=0.50\) in each of two periods, with no other productivity changes. Apply the diffusion equation sequentially to find productivity after two periods. (Answer: period 1 \(=70+0.10(60)=76\); period 2 \(=76+0.10(54)=81.4\))

What You'll Learn