In Avchala, in the middle of nowhere, you might be surprised to find the busy “Craftsmen’s City” or “Khelosnebis Kalaki” as it is called in Georgian. It is built on a 4-ha plot where there used to be a large carpet manufacturing plant in Soviet times, which vanished with the Soviet Union and turned into a concrete carcass surrounded by a swamp.

Luckily, ten years ago, a man called Irakli bought this territory with the intention of renting it out¹ — though it was not an easy task! At a minimum, Irakli had to provide basic conditions for people to move in: improve the access road, bring electricity and gas into the building, fix the roof, etc. After this basic investment, Irakli could attract some clients to his space; however, the rental price was still very low because of the location.

Irakli had to think of something else to increase the value of his space. One option was to somehow increase his future clients’ profitability, which in turn would make them willing to pay a higher price. Inspired by Michael E. Porter’s seminal work, Irakli came up with a brilliant idea. Businesses can benefit from being located close to other businesses working in the same or related industries and form a “cluster”. And because they benefit, they will be willing to pay a higher price for space if there is “synergy” between their business and those of their neighbors: similar manufacturers, providers of inputs, and services.

Now it was time to choose which industry. Irakli understood that his space — far from the city center and not properly renovated — was not a good fit for retail or IT. Based on what he could offer, furniture seemed like a better option. Also, the prospect of local demand for the furniture was bright, due to increasing levels of urbanization, the rise of nuclear families, and the influx of tourists. Moreover, furniture is an imperfectly tradable good² as its transportation costs relative to its value are high, implying that producing and selling locally would be cost-competitive.

It was decided. Irakli would build a furniture cluster! But then came the key question: how to start from zero? Production and investment decisions in the upstream and downstream parts of an industry are often interdependent. When these decisions are made in a decentralized fashion, a new industry may fail to take hold. Furniture producers would not come if there were no input suppliers and service providers, and vice versa, suppliers and service providers would not come if there was no one (or not a sufficient number of manufacturers) to buy their product/service. Irakli had to attract one trailblazer by offering favorable conditions, and guaranteeing that others would follow. In general, there are multiple furniture producers for each supplier in the chain. Thus, attracting suppliers and service providers in the first place by offering them a discount would save Irakli money. So, he let major input suppliers and service providers, not just rent but buy space for a very low price and guaranteed he would bring in furniture manufacturers.

Irakli’s strategy worked very well. Due to increasing demand, rental prices started to rise. The cluster started to function without the need for further intervention on Irakli’s part.

Currently, there are more than 180 companies located in Irakli’s “Khelosnebis Kalaki”, spanning the entire furniture value chain: from suppliers of inputs to providers of services (e.g. design or equipment rental), to furniture manufacturers. Employing a total of close to 2,000 people, these companies differ in size and specialty. It is easy to find almost any necessary expertise within the cluster: wood and steel bending, wood carving, lamination, precision cutting of all relevant materials – natural wood, laminated MDF or plywood panels, glass, marble, and stone.

To sum up, the success was a result of Irakli’s purposeful actions based on two principles: first, there are synergistic benefits of co-location for companies operating in the same industry; second, to start a thing from zero, there is a need for coordination. In other words, geographic concentration offers the possibility of higher productivity, a possibility that can be realized through some kind of coordination.³ In the case of “Khelosnebis Kalaki”, that coordination was conducted in a centralized manner by Irakli himself.

THE GOVERNMENT ACTING LIKE IRAKLI?

Now imagine that the 4-ha plot in Avchala is Georgia, a country that has not been very attractive to invest in so far. Just as furniture producers would not have been brought to Avchala by an “invisible hand”, it is also doubtful that one day, by some mysterious means, a large number of investors will come to Georgia at once and set up a new industry that creates higher value than the present activities in the country and is exportable, thus promising that the country will grow faster.

One might say that it is macroeconomic instability (e.g. exchange rate fluctuations) and weak institutions (in the case of Georgia, geopolitical risks also add up) that hinder poor countries from moving to skill-intensive, high-tech goods. However, as in the case of “Khelosnebis Kalaki”, if the benefits of clustering outweighed the poor location, it may well also help countries such as Georgia to offset those political/economic disadvantages.

Though for clustering to happen, there is a need for coordination and this is especially true when it comes to high-tech products for the following factors:⁴

(1) high-tech production requires a wide range of differentiated intermediate inputs (goods, services, or technologies); more inputs=more players;

(2) some of the intermediate inputs associated with high-tech production are imperfectly tradable; thus, importing is associated with high costs;

(3) these inputs are produced under increasing returns of scale/decreasing marginal costs; thus, to be competitive, there is a need to scale the business.

To be able to produce imperfectly tradable inputs locally, there should be enough demand on the local market to reach economies of scale (in other words, to utilize increasing returns of scale). Alternatively, for the high-tech final-goods sector to become viable, a sufficiently large number of intermediates have to be produced domestically. But if none of these intermediates are currently in production, there may be little incentive for any final-goods firm to enter production on its own. Then, the economy may get stuck in a low-income, low-tech equilibrium.

For example, one imperfectly tradable input is specialized skilled labor, implying that importing it is associated with high costs. To attract investors in electrical equipment manufacturing, for example, there should be electrical engineers on the local market. However, an individual’s decision to specialize in electrical engineering requires that manufacturers already be present and demand the skill. Likewise, schools will not invest in training electrical engineers if there are not sufficient students willing to enter this profession since the investment will not pay off. It is obvious that there is a need for some kind of coordination for a new high-tech industry to take hold.

Another way to think of imperfect traceability of inputs is to consider transport costs and the tacit nature of knowledge.⁶ These two make the cost of importing inputs high. Thus, countries where high-tech production is geographically concentrated gain a competitive advantage. To make co-location happen, there might be a need for coordination in a centralized manner.

What if the Georgian Government were to act like Irakli and serve as a coordinator? Such a strategy is relatively cost-efficient, is faster (than, for example, institution building), and is self-enforcing, as once all the imperfectly tradable input suppliers are present, further intervention from the government is unnecessary. Thus, this is the kind of a role for the government that does not risk it overly interfering in business and further consolidating power.

^{1 This story is based on an interview with Irakli, the founder of “Khelosnebis Kalaki”. For an extended version please see the publication.
2 Good is considered as imperfectly tradable if the costs of trading it with the rest of the world are of importance.
3 See Rodriguez-Clare, A., Rodríguez, F. and Fischer, R., 2005. Coordination failures, clusters, and microeconomic interventions [with Comments]. Economía, 6(1), pp.1-42.
4 See Rodrik, D., 1996. Coordination failures and government policy: A model with applications to East Asia and Eastern Europe. Journal of international economics, 40(1-2), pp.1-22.
5 Ibid.}