Yes, I agree with you that economies based on production of goods in industry or agriculture are less sensitive to external chocs than economies based on services.

How about extraction-based economies?

Do we have actual data?

The answer is not very clear and statistics is needed. Agriculture is very sensitive to weather and food price volatility. Oil producing countries are very sensitive to oil price volatility. Industry is sensitive to scandals (like with Volkswagen in 2015) and trade restrictions. Tourist sector is sensitive to security issues (like restrictions of flights to Egypt by many countries recently).

There was an Economist debate pretty much on this question that you may want to check out. It is online here http://www.columbia.edu/~jb38/papers/pdf/The_Economist_com_Debate_Manufacturing.pdf

Johannes,

Appreciate your link to the Economist debate on Services vs Manufacturing.  Great reading and insights on the pros and cons.

Silburn

Appreciate the feedback and comments Mounir, Doug and Yuri,

@Doug:  you are correct, extractive-industries would be another dimension to be considered

I think this is not correct as services sector is future for job generation. After mechanization there is little scope for generating more jobs in agriculture. manufacturing has also become technology intensive and production can be increased even without much increase in jobs. Services sector on the other hand ranging from car-sales, software, retailing etc can never be replaced through technology. Manufactured goods have become global goods (like I have same S-3 as anyone in developed country with more or less same price) but service is a local phenomenon. For software issue in S-3 I have to rely on local service provider. A professional generates 4 times more jobs than industry.

Appreciate the point Zahid,

In the continuing march of production technology, any human action that is repetitive and routinized will be automated.  This is inevitable and unavoidable.  The impact of this inevitability will have a greater effect on manufacturing (including agro-manufacturing) and agricultural processes than on services processes which are more reliant of human actions.

Silburn

Dear S Clarke, yes.  India's economic growth, recently, based on the growth of its service sector industries more than commodity-sector industries including agriculture and manufacturing industries so India's growth is less sustainable than China because China's economic growth is more based on growth of commodity sector than service sector. That is why Indian Prime Minister Modi initiated 'Making in India' plan and requesting the rich countries, financially and technologically, to start their industries on Indian soil. India, at this juncture, needs a technical breakthrough in commodity sector. I think Clarke may agree my this observation or assumption.

Yes,

Hi all, when one asks a question we need to help him understand. In this context when your answer is yes or no far that matter you need to support it by evidence. In other words you must support your yes or no!

Clearly, absence or low development of industry is bad for any country. Services (apart from financial) have much less possibility to be exported, and domestic demand for services and their price index depends on GDP per capita. Hence, Arjun  is right that India having less share of industry than China has lower possibility of growth. Contrary, growth of service sector in US and UK do not make them poor, also because they were historically industrial countries and now set some production in other countries making only domestic engineering or FDI. This is the first part of robust and sustainable development.

But the question was also about sensitivity to shocks and I mentioned in the first answer that it is not obvious.

The key is to think of an economy like an organism or a patch/niche in a global ecology. Sometimes the situation is such that being too specialized tends to increase the risk of a demise. Sometimes being diversified in the sense of having a bit of everything (sharing aspects that a common to your peers in the ecology), favors your survival. We see this with Iceland and Greece. Which seems to reinforce the believe that service-oriented economies are less resilient. But then we are also entering the era where commodities/tangible goods-based economies are also faltering. So there is essentially no fixed answer to this. In fact if nature and ecology is any to go by, having redundant sectors are vital for resilience.

Not necessarily economies with dominant service sector is less robust, or exposed to losing positions against foreign competition. In this regard I would say that the degree of specialization and complexity of the two sectors is what determines the robusnetss of the economies.

Hi all, even service sectors (close to manufacturing, or finance and IT or Movie industry like Bolly- and Hollywood) can be considerable exporters nowadays, so I do not think one can find a definiteve answer to this question without clearly pinning down and defining what robustness , sustainability and resilience means in operative, indicator terms. Then one should go and dig in the global indicator databases. It could well be that the answer lies e.g. in the educational level of labor force, functioning of the public institutions, flexibility of the labor market etc. I would start with World Bank development indicators.

Dear all, one example of empirical analysis of economic growth is within this book of Barro and Sala-i-Martin from 2004:

http://down.cenet.org.cn/upfile/8/200751171644184.pdf

The current economic and  financial crisis and recovery periods ,at least in the case of Romanian economy, are strong arguments in favor of manufacturing industries or of some agricultural activities as driving forces of crisis exit and recovering growth. 

The decline of services sector  was very slow and  long lasting. May be some services directly related to primary and secondary sectors.could be considered as contributing to recovery. A good part of speculative services sector is very vulnerable to crisis impact and after crisis contribution to growth was very insignificant.   

The knowledge-based reindustrialization of both developed and developing countries at present is a priority of many countries. this could be an argument sustaining manufacturing industry as an important economic sector generating and propagating science and technology in upstream and downstream other economic sectors.

The discussion has been very illuminating.  Many insights gained

Appreciate it all

Silburn

I have read the debate on Services vs Manufacturing that Johanes Hirata recommended in his answer (post number 4). It was really a fierce and interesting debate.

The point of vulnerability to external shock that Mounir Belloumi (1st post) raized and others discussed must also be an important point of discussion. Jouko Antero Kinnunen (post number 15) followed Mounir and others pointing the conceptual ambiguity of robustness, sustainability and resilience.

Zahid Asghar (post number 7) examined if manufacturing can continue to provide a higher percentage of employment.

In his second answer (post number 3) Yuri Yegorov approved Dr. Arjun Yallappa Pangannavar's opinion as follows:

• Arjun is right that India having less share of industry than China has lower possibility of growth.

Is this correct? I do not see any evidence or reasoning. I hope that Pangannavar and Yegorov will expain this point more in detail.

My question here is whether the dichotomy manufacturing versus service is a good classification in discussing robustness and sustainability of economic growth. Many people argued external relations (international trade, intellectual property trade and tourism). Although majority of services are not internationally tradable, some services are "tradable" across countries. When we discuss external factor of economic growth, it would be more plausible to divide all commodities in tradables and intradables. How do you think about it?

Dear Yoshinori,

I think that you agree that historically the period of high growth was linked to industrialization. China had it during the recent decades, Japan and Germany enjoyed high growth after the second world war when their destroyed industry has been rebuilt. USSR had high growth from 1930s till 1960s, again due to industrialization.

Most of services are not tradable, except for financial services and tourism. I have seen statistics that the price index of services has high correlation with GDP per capita. Given that tourism rarely accounts for more than 10-15% of GDP and that there is high global competition to enter such club of countries (Greece, Spain, Austria and only few more are there), Indian perspective of growth without manufacture development might exist, but it will never repeat Chinese speed.

Dear Yuri, 

we must be careful if rapid industrialization caused high growth or high growth rate helped rapid industrialization. 

In the case of high growth period just after the WW II, Japan has already arrived at a certain stage of industrial maturity in 1930’s. Here is the list of percentage of employee (including self-employed) in manufacturing [2nd row] and services (tertiary sector) [3rd row] among all employee in Japan.

1920      1930      1940      1950      1955      1960      1965

16.4       15.9       21.1       15.8       17.5       21.7       24.4

24.2       29.9       29.2       29.7       35.5       38.2       43.7

1970      1975      1980      1985      1990      1995      2000

26.1       24.9       23.7       23.9       23.7       21.1       19.4

46.6       52.0       55.4       57.5       59.4       62.2       65.1

The so called High Growth Period is the two decades from 1955 to 1974. You may say that manufacturing employment increased rapidly, but service sector increased more rapidly. 

In the two decades until the First Oil Shock, real wage rate increased rapidly and it is true that this is supported by high productivity growth. And this productivity growth was stimulated by the increase of production volume which was helped by the avid consumption demand. Real wage growth helped this. In fact, we had a virtuous cycle at that time. Stimulating manufacturing alone cannot realize this cycle. 

Perhaps the same observation may be made for West Germany.

Dear Ivo Carneiro de Sousa

Population and geography are two important conditions when we want to think of economic strategy of an economy. You cited two successful cases i.e. Hon Kong and Singapore. Those economies have a very samll polulation in comparison to China, India and Indonesia. Macao is also a samll economy at the border of big China. Our question setter Silburn Clarke lives in Jamaica. It counts smaller population than Hong Kong and Singapore. In this sense, you are right to cite Hong Kong and Singapore as successful cases.

Comparisons between India and China is interesting in its own right, but they may not be applicable to small economies such as Jamaica and Iceland.

In the case of small economy, it is often difficult to possess various industries that can compete in the global competition. If it can possess only one or two indutries, the vulnerability of the economy depends much on nature of those industries.

Populations (2014, M=million)

Hong Kong　7.2 M　Singapore　5.5 M　Iceland　0.3 M

Macao　0.6 M　Jamaica 2.7 M

China 1,364 M　India 1,295 M　Indonesia 254 M

Dear Yoshinori,

you are right that employment in service sector of Japan was growing along with employment in industry. However, a) the fractions of produced GDP should also be considered, b) there was general global trend in rising share of service employment.

Regarding your comment about small countries, I also agree that they have disadvantage in building multi-sector industry. But they also may have advantage in some special services. Take as examples small European states like Monaco, Luxembourg, Andorra. They either built tax-free shops (Andorra) or set special tax policy that rich people get the residence to pay less taxes and boost housing price (Monaco).

We can make a simple theory for a closed economy on the employment composition among different sectors.

At a time of certain development (thus at a level of per capita income), the total consumption C may be divided into c1C, c2C, and c3C, if we suppose three-sector case. Suppose for simplicity, a good-bundle in each sector is made as a combination of labors in three sectors:

　　v11v12v13 　　for a good-bundle 1

　　v21v22v23 　　for a good-bundle 2

　　v31v32v33 　　for a good-bundle 3

Then employment in each sector is given by

　　(c1v11 + c2v21 + c3v31) C 　　in the 1st sector

　　(c1v12 + c2v22 + c3v32) C 　　in the 2nd sector

　　(c1v13 + c2v23 + c3v33) C 　　in the 3rd sector.

If we write the array of vij matrix V, and ci vector c, employment composition is given by

　　c V.

The matrix V can be calculated by input coefficients A (including labor input coefficients).

In the above, I have neglected investment. This can be done with making a suitable assumption. For the moment, I do not bother with this problem.

Now my point is when two countries are in a similar stage of economic development and their technological knowledge is also similar, then it is highly probable that the two countries have the similar employment composition.

If we want to examine change of employment composition, we have to know the change of matrix V or matrix A. History teaches us that c1 and v11, v21, v31 decline and c3 increases but we do not know well how the other variables change.

For an open small economy, it depends much on international trade and we cannot say anything before we know the composition of the export and import.

What I want to say in this post is that the employment composition of an economy is closely connected to the consumption demand composition via technological coefficients. Actual employment composition (hence proportion between manufacturing and service sector) must be in a good proportion inasmuch as  the economy goes well. 

Although the debate in The Economist was interesting, I do not think that the problem was well posed. Ha-joon Chang opposed manufacturing-based economy against service-based economy. The problem setting has three points of ill-posedness. First, the grouping of industries is too big. The comparison does not apply to small countries. Second, the debate does not make clear distinction between giant countries such as India and China and small countries of various size. Third, real problem is not the internal industrial structure, but the nature of the competitive industry (or industries) of a concerned country.

In this post let me argue the robustness and sustainability problems for (very) small countries.

By the brief examination in above post, it is clear that internal industrial structure cannot be arbitrary without impeding the natural movements of economic flows. It is not wise for a government to intervene into the structure itself. Big difference in industrial structure depends largely on the foreign trade.

Now consider a small country whose population is less than 2 Million but greater than 1 Million . I found 8 such countries: Bahrain 1.4 M, Cyprus 1.2 M, Estonia 1.3 M, Gabon 1.7 M, Latvia 1.99 M, Mauritius 1.3 M, Swaziland 1.3 M, Trinidad and Tobago 1.4 M.

Hong Kong　7.2 M, Singapore　5.5 M, and Jamaica 2.7 M are bigger than these groups in population, whereas Macao　0.6 M and Iceland　0.3 M are smaller than these countries.

I do not know the industrial structure of these countries. If someone know it well, please teach us. It would be difficult for these countries to develop their economy as a closed economy. They must have important industries by which they trade with the world. However, it would be also difficult to have many such industries. Suppose they can have only one such industry and call such industry the competitive industry of the nation.

Now the robustness & sustainability of a small country would depend largely on the nature of the competitive industry.

Silburn Clarke raised, as examples of possibly competitive service industries, finance, tourism, and ICT. Jouko Antero Kinnunen gave an examples of movie industry. As for manufacturing, I do not know good examples. In the case of Estonia, its competitive industry seems to be wood processing. As for agriculture, we may give industries such as sugar, coffee, banana and others. As for mining, we many give oil, rare metal and others.

We now know that the tourism is quite vulnerable to events like terrors. Agricultural products depend on climate and on external market. Mining has the problem of exhaustion. In view of robustness & sustainability, we cannot say such and such industry is a good selection. To depend on any one industry may not be a good strategy for a country. Then, what I can imagine as a possible solution is to enhance capability to adapt to the new situation. How do you think about it?

Dear Yoshinori,

Appreciate the generous feedback.

Agree the point that an examination of the contribution of the tradeable versus nontradeable components of the national economy, whether services or manufacturing,  may yield better insights on robustness, resilience and sustainability of these economies.

Would you have a disaggregation of GDP data by tradeable vs nontradeable size and sectors for your selected group of 13 small countries (Bahrain 1.4 M, Cyprus 1.2 M, Estonia 1.3 M, Gabon 1.7 M, Latvia 1.99 M, Mauritius 1.3 M, Swaziland 1.3 M, Trinidad and Tobago 1.4 M., Hong Kong 7.2 M, Singapore　5.5 M, Jamaica 2.7 M,  Macao　0.6 M and Iceland 0.3 M.  As a comparator for the group, I would also add the Uruguay 3.4M and the OECD member Finland  5.4 million.

It would be interesting to see the pattern of this data.

Also agree with the assertion that small countries have to operate open economies.  However by sheer limitation of size, these countries would have constrained capabilities (talent, intellectual capital, R&D), which would tend to limit the number, ie range, of industries possible, to a few and consequently invite increased vulnerabilities.  Is it therefore the fate of these economies to be vulnerable ?  Or does participation / interdependence in regional blocs, eg ASEAN, help to mitigate the risks and vulnerabilities of a single open-economy approach ?

Silburn

Dear Ivo,

Appreciate the responses.

The description of the various pull factors operating through globally integrated networks brings to mind the Flying Geese Paradigm originated by Kaname Akamatsu in the 1930s.

It would suggest that small economies needs the locomotive effect of these leading geese to pull them along

Silburn

Dear Silburn,

If you have disaggregated GDP data by tradeable vs nontradeable size and sectors for the group of 13 small countries, please let me know them. If we all consider such data, we may find a good idea about the future strategy of those countries.

You have mentioned Flying Geese theory of industrial development. Akamatsu's Flying Geese pattern is now famous because it partly explained East Asian development. However, his theory comprises three different models, which he has detected in different steps. What is famous as his FG pattern is the second one. East Asian countries seemed sometime to make a Flying Geese formation with Japan at the top and other countries following it. In fact, this is Akamatsu's second model of the FG theory.

The first model is the pattern that we often observes for a country. For example, Japan first imported woolen textile from abroad.  When its volume became large, a Japanese firms started to make wool protected by differential importation tariffs (low or no tariff for importation of law materials and high tariff for finished products). When the production experience accumulates, productivity becomes higher and exportation becomes possible in the international market. The important factor in this transition from importation, import substitution and export is low wages with regard to developed countries. For the details, please see the paper below in RG.

The second model is famous one and I can skip it.

The third model concerns the world FG structure. Key words are homogenization and heterogenization. If top leading countries continue to produce new products or services, world-wide FG pattern is preserved. If top countries decelerate the speed of new product development, then the catching up proceed and the world structure will approach to a more homogenized patter.

As you see at the first model, importation of raw materials, ingredients, and parts are crucial part of the story. However, in trade theory, this importation of raw materials, ingredients, and parts were not incorporated into trade theory for a long time. They are now called input trade. In East Asian tradition, export of processed foods of imported materials is called processing trade. This was indeed a key factor even for the British Industrial Revolution. English language lacked even a concept for it. This is an enigma I am now attacking. Please look at  my question:

Why did the notion of "processing trade" not become a major concept before fragmentation and global supply chain? https://www.researchgate.net/post/Why_did_the_notion_of_processing_trade_not_become_a_major_concept_before_fragmentation_and_global_supply_chain

Conference Paper The Economics of the Great Unbundling