A central point —emphasized by development economists since Smith (2002) first wrote about the employability of labour by capital —is that productive capacity is a function of the amount of capital, which enhances labour productivity. Consequently, the same economic activity can yield higher output (for empirical evidence, see Kaldor (1957, 1961)). Moreover, because labour is employed through capital, the level of employment is directly linked to capital accumulation which stems from investment. Tracking investment is therefore crucial for analysing economic growth, as it determines the capital stock.

The neoclassical notion of capital can be broadened beyond physical durable goods to include the knowledge and skills acquired by workers over time (Barro, 1996). However, for the sake of simplicity and for the purpose of this article, the term will be used to refer exclusively to physical capital.

As shown in Figure 3, there is a striking resemblance between investment behaviour and the evolution of capital-goods imports for agriculture and manufacturing. On one hand, the upward trend in investment aligns with the increasing pattern of capital-goods imports. On the other hand, investment tapers off when there is a significant decline in such imports. This is unsurprising given Colombia's reliance on imported capital goods — such as machinery and equipment — a common feature of resource-rich economies (Smith, 2007).

Figure 3. Evolvement of investment and capital goods imports.

It can be argued that given the Colombian economy's reliance on capital-goods imports, it is essential to maintain equilibrium in the external sector —measured by net exports —whenever investment accelerates. To achieve this, exports must be vigorous enough to sustain high investment rates without compromising external balance (Thirlwall, 2012).

When the economy liberalized in 1991-1992, both investment and imports increased significantly—by 47.7% and 39.4%, respectively—and the correlation between these two variables strengthened. Since then, the share of imports and exports in GDP (economic openness) has risen steadily. These stylized facts suggest that investment rates have become increasingly dependent on imports as international trade gained relevance in Colombia's economy. Figure 4.

Figure 4. Investment- imports rolling correlation of 10 years and Economic openness.

Rolling correlation of 10 years is measured in the left axis while the economic openness is measured in the right axis.

The scatterplot presented in Figure 5 is based on data from 2001 to 2021, comprising 21 data points. It illustrates Colombia's dependence on foreign capital goods, with higher investment rates corresponding to increased imports. Although a larger sample size would be required to draw definitive conclusions, the positive correlation between the variables and the R² value of 0.44 are noteworthy findings —particularly given that variations in capital-goods imports are explained by a single variable: investment.

Figure 5. Correlation between Investment and Capital Goods Imports, 2001-2021.

It is noteworthy to mention that DANE’s data of capital goods imports is made up of both agriculture and manufacturing capital goods.

When examining the behaviour of imports by type of capital goods, it is evident that imports od capital goods for the manufacturing sector exhibit a stronger correlation with investment than those for agriculture sector. Figure 6.

Figure 6. Correlation between investment and Capital goods imports by sector.

Imports of agricultural capital goods showed no correlation with investment between 2001 and 2021. This finding suggests that Colombia's manufacturing sector relied heavily on foreign capital goods, while the agricultural sector depend more on domestically produced capital goods. In some cases, farmers may have been able to substitute imported technology with local alternatives —for example, opting for organic rather than synthetic fertilizers. However, it is important to note that this article does not aim to extensively examine the agricultural sector in detail.

The pattern observed for imports of raw materials and intermediate goods follows a trend similar to that capital goods. The manufacturing sector exhibits greater dependence on foreign production compared than agriculture. In contrast to the capital-goods case, however, the agricultural sector shows a higher correlation. This suggests that Colombia's agricultural sector relied more on foreign-sourced raw materials and intermediate goods than on foreign capital goods. Figure 7.

Figure 7. Correlation between investment and raw materials and intermediate goods imports by sector.

The correlations discussed above highlight the relationship between investment and imports according to their respective types. Manufacturing demonstrates greater reliance on capital-goods imports than imports of raw materials and intermediate goods, as reflected in the R² values from each regression. This interpretation aligns with Prebisch's (1988) broader notion of the dependence of peripheral economies on industrial centers.

The following table presents the elasticity of imports with respect to GDP and investment. Overall, the evidence shows that imports are responsive to changes in GDP, meaning that the expansion of Colombia’s economy has led to increased imports. By contrast, the elasticity of imports with respect to investment displays greater variability, shifting between elastic and inelastic over time. Notably, between 2002 and 2015, the investment elasticity of imports followed an inverted U-shaped pattern, peaking at 7.12 in 2009. These years of elastic imports align with the resource boom period, as illustrated below. Table 4.

Table 4. Elasticity of imports to GDP and investment.

Next, we analyze Colombia's integration into the world economy and the behavior of the exchange rate.

In 1992 Colombia implemented a set of reforms to its economic model, adopting a new approach based on the theory of comparative advantage (Document CONPES 2465, 1990). These reforms aimed to expand the role of market forces by dismantling measures such as Decree-Law 444 of 1967 (Decreto 444, 1967), which had focused on export promotion and diversification. As a result, the market began to play a more prominent role in allocating resources across industries. It is important to note that these changes occurred within the broader context of global neoliberal policies, which, as described by Palma (2019), led to a prolonged process of deindustrialization.

An examination of Colombia’s exports reveals that the mining sector experienced significant growth in its share between 2004 and 2014, reaching nearly 60% of total exports — a period that corresponds to the resource boom, discussed later. Table 5 presents the shares of the three major sectors — agriculture, mining, and manufacturing — in total exports. From 2010 to 2018, Colombia’s export structure was predominantly specialized in mining activities, reflecting a strong orientation toward resource-intensive industries.

Table 5. Exports by Sector as a percentage of total exports.

When examining the resource sector, it becomes clear that oil and coal are the dominant subsectors, together accounting for an average of 42% of total exports between 2000 and 2020. The peak occurred in 2014, when oil and coal reached a remarkable 60% share of total exports. Thereafter, their share gradually decline, eventually falling to 36% (Figure 8). This raises two critical questions: Why were these two export activities unable to sustain their growth and instead experienced a slowdown? And Does the observed behavior of oil and coal exports reflect an endogenous or exogenous trend? Figure 7.

Figure 7. Exports of land-based primary goods.

By tracking the interplay between international prices and the exports of oil and coal, Figure 9 highlights the resource sector's strong dependence on global price trends. The data reveal a resource-price boom swiftly followed by a decline in international prices for Colombia’s primary exports activities —oil and coal. (Calculations are based on annual average, considering the mean of both price indexes from the year the decline began through 2016). Oil prices were measured using the Brent and West Texas indexes, while coal prices were based on the Northwest Europe and US Central Appalachian indexes. To smooth representation of the data, a five-year moving average was applied to the international price series.

Figure 9 clearly captures the oil boom, which began in 2000 and peaked in 2014. In contrast, Figure 10 shows the coal boom starting in 2000 and ending in 2012. These findings underscore the resource sector's vulnerability to international price dynamics, as both oil and coal experienced boom periods that subsequently waned due to global market fluctuations.

When the same exercise is applied to the coal industry and its international prices, the relationship becomes evident. Figure 11 demonstrates that the proportion of coal in exports aligns with the fluctuation of international prices. Figure 8.

Figure 8. Total and mining exports and International Prices of Oil.

On the other hand, foreign direct investment (FDI) has been a significant source of foreign currency. During the resource boom, there was a substantial increase in foreign capital, primarily directed towards the resource and service sectors. However, once the upswing ended, average FDI levels declined. Notably, mining and services emerged as the two most dynamic sectors, experiencing growth between 2000 and 2004, peaking from 2010 to 2014, and subsequently witnessing a significant decline (Figure 9).

Of particular interest is the sharp downturn in the manufacturing sector between 2020 and 2022, a decline largely attributable to lockdown policies in response implemented to the COVID-19 pandemic.

Figure 9. Flows of Foreign Direct Investment 2000-2021.

Similar to exports, foreign direct investment (FDI) also appears to have been influenced by international commodity prices. As shown in Figure 10, FDI closely followed the trend of oil and coal prices. However, when prices began to decline gradually yet substantially, investment in the mining sector fell sharply, followed by subsequent drop in FDI directed to the services sector. This pattern suggests that falling international commodity prices discouraged foreign capital investment in mining.

Figure 10. Oil FDI in USD Million, International Prices of Oil and Coal in Dollars per Unit.

The inflow of foreign exchange slowed as both export performance weakened and FDI declined. Consequently, imports could no longer continue expanding, and an adjustment had to take place by diminishing its pace. In other words, as the inflow of foreign exchange decrease, imports of capital goods and investment must slow down –an outcome predicted by Thirlwall’s (2012) model of the foreign exchange constraint. Figure 11 illustrates the piling up of international reserves during the resource-boom, and the slowdown of its growth rates once such as the commodity boom came to an end in 2014.

Figure 11. International Reserves and commodity-international prices.

Figure 12 displays the interaction between the growth of international reserves and the value of the Colombian peso (COP) against the US dollar (USD). Between 2000 and 2014 — a period that encompasses the resource boom — international reserves accumulated, and the COP appreciated against the USD. However, once coal and oil prices declined, the growth of reserves slowed significantly, and the COP depreciated by 63%, reaching 3,054 pesos per dollar, an exchange rate previously unprecedented in the country’s history. It appears that after the resource boom came to an end, no other sector of the economy was able to generate the same level of international reserves as the resource-based industries (oil and coal).

Figure 12. International Reserves and Exchange rate.

The economy measured by the Real GDP evidently was affected by the commodities prices and its growth rates fluctuated accordingly with the fluctuation of international prices of both oil and coal. Figure 13.

Figure 13. Real GDP Growth, Oil and Coal International Prices.

Looking at the trends, when international prices were rising economic growth rates also increased. However, once prices peaked and then declined, the economy slowed down; the dependency of the whole economy on resource-booms is patent. The external sector strongly depends on oil and coal production, such industries that are highly affected by international prices.

Table 6 presents available data on Gross Fixed Capital Formation by component. Although the series extends only to 2016, it nonetheless allows for an analysis of the resource-boom period. It is noticeable that when oil and coal prices soared, the investment grew faster in both items 1) transport equipment and 2) machinery and equipment, which makes sense keeping in mind that the growing products turned out must be transported (e.g. oil barrels and coal tons) and the investment in physical capital (machinery and equipment) could be carried out while the economic activity was on the rise.

Nevertheless, during the post-boom downswing, investment in those items fell significantly, declining by -1.5% in machinery and equipment, and -7.2% in transport equipment. The investment was redirected toward nontradable internal activities, namely civil works and building construction. It may be a case of Dutch disease effect, as analyzed by Brahmbhatt, et al. (2010). Overall, however, the investment decreased.

Table 6. GFCF share by sector and International Prices of Commodities changes.

Up to this point, it is evident that Colombia, as a small open economy heavily reliant on exporting resource commodities under a free-market policy orientation, appears to be compelled to deepen its focus on resource-driven trade practices. In this context, currency devaluation is often considered a viable strategy to strengthen the external economic sector, as discussed in Guerron's (2013) work. In the following section, we will delve deeper into this subject.

The following scatterplot shows the correlation between imports of capital goods and the depreciation of the Colombian peso against the US dollar, measured by the percentage change in the percentage change in the COP/USD exchange rate. The regression line indicates a negative relationship between these two variables, with a R² value = 0.4416. In other words, that is to say, when the Colombian currency ramped up in value, total imports rose, and vice versa, when the Colombian peso depreciated, total imports dropped. Figure 14.

Figure 14. Relationship between total Imports and depreciation of domestic currency 2000-2021.

The negative relationship is higher between imports of capital goods and the depreciation of the domestic currency (R² = 0.4796). It is an indication of foreign exchange constraint that comes about with the weakness of domestic currency, which negatively affects investment through the increase in the price of capital goods –a case of exchange rate pass-through. The relationship is even higher if we observe how capital goods imports for manufacturing and exchange rate correlate (R² = 0.4812). If we plot the correlation of exchange rate with capital goods imports for agriculture, R^2 is not significant (R² = 0.0861), showing a weak relationship between the two variables. Figure 15.

Figure 15. Relationship between imports of capital goods for manufacturing and exchange rate.

We can assert that Colombia’s manufacturing sector depends on capital-goods imports and is arguably negatively affected by exchange rate increases (the depreciation of the Colombian peso against the US dollar), which make industrial inputs more expensive. From 2012 to 2021, the exchange rate rose, weakening the value of the domestic currency and driving up the prices of imported inputs demanded by producers. The Producer Price Index (PPI) for imports followed a similar pattern, displaying a U-shaped trajectory alongside the exchange rate. Furthermore, the outbreak of the COVID-19 pandemic between 2020 and 2021 contributed to the rise in import prices. Figure 16.

Figure 16. Producer price index of imports and exchange rate.

Technological dependence remains a key concern for Colombia’s economic growth (Ortiz Quevedo & Vásquez Castro, 2007). The capacity to import has long been recognized as a structural hurdle in Latin American economies, while manufacturing is viewed as a crucial vehicle for economic development (Prebisch, 1988). From the perspective of Latin American structuralism, balance of payments constitutes a key factor to be solved in order to hold high rates of economic growth (Ocampo, 2014).

What we have seen throughout the stylized facts stood out here endorses the concern on external constraints in the economic development of Colombia and provides insights for further discussions regarding the constraints set up by the amount of foreign exchange in the case of Colombia. These facts also are important for considerations related to the use of international reserves to boost economic development –for instance Gonzales (2016) argues for the potential use of foreign exchange to finance public spending and the optimal level of reserves.