Why Technology Access Is Not the Same as Technology Use
A firm or economy can have access to a technology without being able to use it productively. A new production process may require trained workers, reliable infrastructure, effective management, and enough existing knowledge to understand how the process works. If these complementary capabilities are weak, the technology may be adopted slowly, used inconsistently, or fail to deliver its potential productivity benefit.
Absorptive capacity describes the ability to recognize the value of external knowledge, understand it, adapt it to local conditions, and put it to productive use. It is a capability, not the technology itself and not a measure of how much technology is available. The distinction matters for productivity catch-up: two economies exposed to the same external technology may benefit at different rates because they differ in their ability to absorb it.
Four commonly important contributors are workforce skills, management quality, infrastructure, and existing knowledge. Skills help workers operate unfamiliar equipment and learn new procedures. Management supports coordination, training, and changes to production routines. Infrastructure such as reliable power and communications makes it possible to deploy technology consistently. Existing knowledge provides a foundation for recognizing, evaluating, and adapting unfamiliar ideas.
These contributors may reinforce one another. Skilled employees cannot fully exploit a process if power supply is unreliable; good infrastructure alone does not ensure that managers reorganize production appropriately. A simple index is useful for transparent calculations, but it should not be mistaken for a claim that capabilities are always independent or perfectly substitutable.
A Simple Model of Effective Adoption
Let \(A_t\) denote local productivity at time \(t\), and let \(A_F\) denote the productivity associated with an available external technology or frontier. For the calculation below, assume \(A_F>A_t\). Let \(\lambda\) be the baseline rate at which the external technology is available for potential adoption, with \(0\leq\lambda\leq1\). Let \(C_t\), between 0 and 1, represent absorptive capacity. The model says that capacity scales the share of the remaining productivity gap closed in one period:
The product \(\lambda C_t\) is the effective adoption fraction for the period. It is not the capacity index alone. For example, a capacity index of 0.7 combined with a baseline adoption rate of 0.4 implies an effective fraction of 0.28, not 0.7. The equation is a reduced-form representation: it does not specify how skills or infrastructure are produced, nor does it describe research that creates a new frontier.
The equation is most appropriate when the frontier is treated as available, productivity is measured on a consistent scale, and the period’s effective adoption fraction does not exceed 1. The simple interpretation also holds \(A_F\), \(\lambda\), and \(C_t\) fixed during the period. If adoption is unavailable, \(\lambda=0\); if effective capacity is zero, \(C_t=0\). Either case implies no productivity gain from this channel. If the frontier changes over time or capacity changes during adoption, the calculation must be updated period by period.
Worked Example: Calculating an Effective Adoption Fraction
A hypothetical economy assesses four normalized capabilities, each on a scale from 0 to 1. Its skills score is 0.8, management score is 0.6, infrastructure score is 0.5, and existing-knowledge score is 0.9. A planning model assigns respective weights of 0.40, 0.25, 0.20, and 0.15. The external technology has a baseline adoption rate \(\lambda=0.40\). Local productivity is \(A_t=100\), and frontier productivity is \(A_F=150\). Calculate the capacity index, the productivity gain, and next-period productivity.
Method. Calculate the weighted capacity index, then multiply it by the baseline rate to find the effective adoption fraction. Apply that fraction to the productivity gap.
Capacity index. The weighted contributions are \(0.40(0.8)=0.32\), \(0.25(0.6)=0.15\), \(0.20(0.5)=0.10\), and \(0.15(0.9)=0.135\). Therefore,
Effective adoption fraction. The capacity-adjusted fraction is \(\lambda C_t=0.40(0.705)=0.282\), or 28.2% of the remaining gap.
Productivity gap and gain. The gap is \(A_F-A_t=150-100=50\). The gain is \(0.282(50)=14.1\), so \(A_{t+1}=100+14.1=114.1\).
Check. The remaining gap is \(150-114.1=35.9\). Also, \(0.718(50)=35.9\), because \(1-0.282=0.718\). The values are consistent. The common mistake this method avoids is treating the capacity index, 0.705, as the adoption fraction without multiplying by \(\lambda\).
Why the Gap-Closes Formula Holds
The adoption equation can be understood by tracking the part of the productivity gap that remains after one period. Define the gap as \(G_t=A_F-A_t\). Under the model, the economy gains \(\lambda C_tG_t\) in productivity. Subtract that gain from the old gap:
\(G_{t+1}=A_F-A_{t+1}\). This is the difference between the unchanged frontier and next-period local productivity.
\(G_{t+1}=A_F-[A_t+\lambda C_t(A_F-A_t)]\).
Since \(A_F-A_t=G_t\), \(G_{t+1}=G_t-\lambda C_tG_t=(1-\lambda C_t)G_t\). Thus the remaining gap is the old gap multiplied by one minus the effective adoption fraction.
If \(A_F\), \(\lambda\), and \(C\) are constant over \(n\) periods, applying this relation repeatedly gives \(G_{t+n}=(1-\lambda C)^nG_t\). The accumulated productivity gain is therefore \(G_t-G_{t+n}\). This geometric result is conditional on the constant-parameter assumptions; it should not be used unchanged if capacity or the frontier varies across periods.
Worked Example: Comparing Catch-Up Across Two Economies
Two hypothetical economies start at productivity \(A_0=40\), while the available frontier is \(A_F=100\). Both face the same baseline adoption rate, \(\lambda=0.40\), and their capacity scores are held constant for two periods. Economy X has normalized scores \(H=0.9\), \(M=0.4\), \(I=0.8\), and \(K=0.5\). Economy Y has scores \(H=0.6\), \(M=0.8\), \(I=0.5\), and \(K=0.8\). Use weights \(w_H=0.40\), \(w_M=0.25\), \(w_I=0.20\), and \(w_K=0.15\). Which economy has higher productivity after two periods?
Method. Calculate each weighted capacity index, then apply the gap equation for two periods. Both economies begin with a gap of \(100-40=60\).
Economy X. Its capacity index is \(0.40(0.9)+0.25(0.4)+0.20(0.8)+0.15(0.5)=0.36+0.10+0.16+0.075=0.695\). Its effective adoption fraction is \(0.40(0.695)=0.278\). The remaining-gap factor per period is \(1-0.278=0.722\). After two periods, the gap is \(60(0.722)^2=60(0.521284)=31.27704\). Therefore, productivity is \(100-31.27704=68.72296\), or 68.72 to two decimal places.
Economy Y. Its capacity index is \(0.40(0.6)+0.25(0.8)+0.20(0.5)+0.15(0.8)=0.24+0.20+0.10+0.12=0.66\). Its effective adoption fraction is \(0.40(0.66)=0.264\), leaving a per-period gap factor of \(0.736\). After two periods, the gap is \(60(0.736)^2=60(0.541696)=32.50176\). Productivity is \(100-32.50176=67.49824\), or 67.50 to two decimal places.
Conclusion and check. Economy X has productivity 68.72, above Economy Y’s 67.50. X’s remaining gap, 31.27704, is smaller than Y’s 32.50176, as expected from X’s higher capacity index and adoption fraction. The common mistake avoided here is comparing only one period or comparing the capacity scores without applying them to the remaining gap over both periods.
Capacity Components and Bottlenecks
A weighted index makes the contribution of each capability explicit. For normalized scores \(H\), \(M\), \(I\), and \(K\), one illustrative specification is:
The weights are model choices that should be justified for the application; they are not universal constants. If all four scores are between 0 and 1 and the weights are nonnegative and sum to 1, then \(C\) is also between 0 and 1. A higher index increases the effective adoption fraction in this particular model, all else equal. It does not establish that every capability has the same causal importance in every setting.
The additive specification also allows one strong component to offset a weak one. That may be an unrealistic assumption when capabilities are complements. If technology needs both trained operators and reliable power, a severe shortfall in either may constrain adoption more than a weighted average suggests. A practical interpretation is to examine both the summary index and the underlying scores, rather than treating a single number as a complete diagnosis.
Worked Example: Assessing an Infrastructure Improvement
A hypothetical economy uses the same weighted index, with weights \(0.40\) for skills, \(0.25\) for management, \(0.20\) for infrastructure, and \(0.15\) for existing knowledge. Its scores are \(H=0.5\), \(M=0.7\), \(I=0.4\), and \(K=0.6\). Consider a scenario in which the infrastructure score rises to 0.9, while the other scores remain unchanged. The baseline adoption rate is \(\lambda=0.50\), and the productivity gap is 80. Compare the model-implied one-period gains before and after the change.
Method. Recalculate the capacity index, multiply by the same \(\lambda\), and apply each effective adoption fraction to the same starting gap.
Before the change. The capacity index is \(0.40(0.5)+0.25(0.7)+0.20(0.4)+0.15(0.6)=0.20+0.175+0.08+0.09=0.545\). The effective adoption fraction is \(0.50(0.545)=0.2725\). The one-period productivity gain is \(0.2725(80)=21.8\).
After the change. The infrastructure contribution rises from \(0.20(0.4)=0.08\) to \(0.20(0.9)=0.18\), an increase of 0.10 in the index. The new index is \(0.545+0.10=0.645\). The new effective adoption fraction is \(0.50(0.645)=0.3225\), and the gain is \(0.3225(80)=25.8\).
Comparison and check. The modeled gain rises by \(25.8-21.8=4.0\). The effective adoption fraction rises by 0.05, and \(0.05(80)=4.0\), confirming the difference. This is a conditional model comparison, not proof that a real infrastructure change would cause exactly this productivity increase. The common mistake avoided is applying the 0.10 increase in the capacity index directly to the gap: the baseline rate \(\lambda=0.50\) must also be applied.
Interpreting the Results
Absorptive capacity helps explain why technology diffusion and international catch-up are not automatic. A larger gap from the frontier creates a larger potential gain for any given adoption fraction, but it does not by itself raise that fraction. In the model, a larger \(A_F-A_t\) raises the gain at a fixed \(\lambda C_t\); a smaller \(C_t\) lowers the share of that gap closed. Thus, a substantial technology gap can coexist with slow catch-up when capabilities are limited.
When productivity is measured so that output is proportional to productivity with other inputs held constant, a productivity increase raises output in the same proportion. That conditional statement isolates the productivity channel; it does not account for changes in labor, capital, or other determinants of output. In applied analysis, the assumptions behind any capacity index and adoption rate should be stated, and results should be read as model-based comparisons rather than precise forecasts.
This topic is aligned with the SOA VEE Economics (Validation by Educational Experience) syllabus. For exam-style calculations, keep three quantities distinct: the capability index \(C\), the effective adoption fraction \(\lambda C\), and the productivity gap \(A_F-A_t\). Their product determines the period’s productivity gain under the model.
Check Your Understanding
For each question, use the stated model and assumptions; answers are given at the end of each item for checking.
- A hypothetical economy has normalized scores \(H=0.7\), \(M=0.5\), \(I=0.8\), and \(K=0.6\), with weights \(0.50\), \(0.20\), \(0.20\), and \(0.10\), respectively. Its baseline adoption rate is \(\lambda=0.30\). Find its capacity index and effective adoption fraction. (Answer: \(C=0.67\); effective fraction \(=0.201\), or 20.1%.)
- Local productivity is 60, frontier productivity is 100, capacity is fixed at \(C=0.40\), and \(\lambda=0.50\). Calculate next-period productivity and the remaining gap. (Answer: \(A_1=68\); remaining gap \(=32\).)
- A region begins at productivity 30, with a fixed frontier of 90, \(C=0.80\), and \(\lambda=0.25\). Find productivity after two periods, assuming capacity and the frontier remain constant. (Answer: \(A_2=51.6\).)
- Use weights \(0.30\), \(0.30\), \(0.20\), and \(0.20\) for skills, management, infrastructure, and existing knowledge. Scores are \(H=0.5\), \(M=0.6\), \(I=0.4\), and \(K=0.7\). Management rises from 0.6 to 0.9, with other scores unchanged. If \(\lambda=0.50\) and the initial gap is 50, compare the one-period gains. (Answer: capacity rises from 0.55 to 0.64; gains rise from 13.75 to 16.00, an increase of 2.25.)
- An economy has a productivity gap of 40, capacity \(C=0.75\), and baseline adoption rate \(\lambda=0.20\). A calculation reports a gain of 30 by multiplying the gap by capacity alone. Find the model-implied gain and identify the error. (Answer: the gain is \(0.20(0.75)(40)=6\); the incorrect calculation omitted \(\lambda\).)