Cover Image
close this bookEmigration Pressures and Structural Change. Case Study of the Philippines (International Labour Organization, 1997, 56 p.)
View the document(introduction...)
View the documentForeword
View the document1. Introduction
View the document2. Migration and the unemployment problem
View the document3. Migration pressures: Supply-side push factors
View the document4. Migration pressures: Demand-side pull factors
View the document5. Migration: Supply-driven or demand-induced?
Open this folder and view contents6. Dimensions of labour emigration
Open this folder and view contents7. The poverty nexus: Dual migration circuits
Open this folder and view contents8. A strategy of selective interventions
View the documentBibliography
View the documentInternational Migration Papers - Cahiers de Migrations Internationales - Estudios Sobre Migraciones Internacionales

3. Migration pressures: Supply-side push factors

As far as economic migration is concerned, two contending views exist: one would assert that large scale international migration is essentially demand induced; the other would insist that it arises instead from unbearable supply-side pressures in the sending economy. Causality is thus sourced in either the labour importing or in the labour exporting economy. Had there been no restrictions on the movement of labour, and should such movement have been relatively inexpensive and well-informed, such pressures could not have built up in the first place. However, the application of restrictions allows the intended market-induced flows to be dammed and thus to build up pressure on the withholding system. But how acute is such pressure? This paper examines the two elements, and subsequently attempts a resolution. The argument is developed explicitly with regard to the Philippines.

In conceptualising supply-side push forces, it would be useful to distinguish between three aspects: first, structural dimensions which have the character of “stock” levels; second, flow indicators pertaining to the direction and rate of change of key indices which explain excess labour supply; and third, the nature of expectations of the future performance of key variables. Evidence from a wide range of relevant indicators is assembled below, and some processal and causal connections are identified. The story is a rather dismal one.

As usual, periodisation with respect to macro-economic performance poses a few controversial choices (Data used are drawn from the Philippine Statistical Yearbook, 1994 and earlier issues). An optimistic way of telling the story would say: GDP grew annually at over 5 per cent between 1975 and 1982; dropped by about 15 per cent over the next two crisis years; grew again at about 4.5 per cent per annum between 1985 and 1990; and after a few hiccups, seems to have resumed this rate since 1993. But the cup certainly was (at least) half empty, and a different rendition, using a slightly different periodisation, would point out that the economy grew between 1982 and 1992 at under 1 per cent per annum. This has to be set against a population growth rate of 2.3 per cent per year; per capita GDP fell at 1.4 per cent per year, and was about 15 per cent lower in 1992 than in 1982. The 1980s could be labelled the Philippines’ lost decade, except that this was when the political transition took place. It is also arguable that the stagnation and decline have much to do with the past regime’s structures and policies. Some credence is provided by the first periodisation, where the villain of the piece is the crisis of the mid-1980s. However, the current signs of recovery will need to be sustained for longer before they can be described as a sustained recovery. Should this happen, indeed, the stagnation of the 1980s could be added to the balance sheet of the old regime; if the recovery remains fitful and obstinate in the future, the more structural and external conditions of the economy would be highlighted. Speculative arguments over statistical interpretation cannot, however, make the stagnation of the 1982-1992 period disappear!

On the side of labour, the starting point must be the population growth rate. At under 2.5 per cent per annum, it cannot be called “high,” except that countries with the levels of income, urbanisation, female educational attainment and female labour force participation rates could have been expected to have moved to a rather lower growth rate by now. In part, this may be attributed to cultural and religious factors; in part, the effect of a past surge in fertility accounts for a ripple effect on the growth rate.2 Before pinning the blame on this factor, it should be remembered that through much of the initial growth drive of the Republic of Korea and Taiwan (China), their population growth rates were much higher.

2 Proponents of strong population control policies could, as customary, argue that high population growth rates were dysfunctional in terms of GDP growth, and hence a lower population growth rate would have transmitted itself to per capita income growth. On the other side, arguments over free choice, respect for cultural and religious norms would no doubt be invoked by another kind of conservative opinion.

Against the population growth rate of 2.3 per cent, the labour force, i.e. population aged fifteen years or more, grew much faster at 2.9 per cent per year. Compounding this was a noticeable rise in the overall labour force participation rate, from 60.1 per cent in 1982 to 65.0 in 1992; this increase was more pronounced in the case of women. This implies an annual growth rate of the active labour force of 3.5 per cent per year for the reference decade. Should a “normal” figure be assumed for the elasticity of employment with respect to output, say 0.5, this would call for a growth rate of 7 per cent per annum. With stagnation, employment levels should have suffered as well.

In reality, employment grew even faster than the growth rate of the active labour force, i.e., at 3.6 per cent per annum which, given the 1.0 per cent per annum growth rate of GDP, yields a nonsensical actual figure for this elasticity.3 Not surprisingly, aggregate productivity levels plummeted, falling at 2.6 per cent per year to a little over three-quarters the starting point. Alongside this, the rate of unemployment actually fell from 9.4 per cent to 8.6 per cent, and the rate of underemployment from 25.5 per cent to 19.9 per cent. The data seem to suggest that the average Filipino “worked” more in order to retain his or her claim to a shrinking pie. The increased labour participation and employment indices probably reflect the outcome of intense labour market survival strategies.4

3 The interpretation of such elasticities should, but seldom does, take changing levels of productivity into account.

4 Being recorded as “employed” might hide the drop in the amount of work being done per person. There is a limit to labour absorption, even in self-exploitation activities of a Chayanovian character. There is some survey evidence that this indeed was happening in rural Philippines. Sanchez (1994:21-22) cites research which shows that the average hours of work in agriculture during the period 1971-1985 declined from 42.1 hours per week in 1971 to 32.8 hours in 1980 and then to 25.6 hours per week in 1985, a staggering drop of 40 per cent.

Though the data for the incidence of poverty and the degree of income inequality are available for somewhat different benchmark years, and with the usual range of qualifications regarding comparability, it is arguable that the reference period saw a reduction in headcount poverty as well as inequality (Balisacan, 1991: Tables 4,5; Intal, 1994: Tables 1,2). If this is valid for the reference years as well, it would imply a parallel process of the sharing out of poverty; the fall in aggregate product would have fallen disproportionately on the upper income classes.

There can be little doubt that all this would have intensified labour market crowding. Real wages for Manila unskilled labour rose marginally by 0.8 per cent per annum. Rather against the trend, however, real wages for farm workers rose sharply by 3 per cent per annum, a puzzle that needs to be resolved.

While the share of agriculture in GDP remained stationary at 22.9 per cent between the two years, implying an identical growth rate for the agricultural component of the GDP, there was much turmoil beneath this deceptively becalmed surface. In contrast, agricultural GDP per person employed in the rural sector rose by 0.1 per cent per year (as against a drop of 2.6 per cent per year for all GDP including agriculture); this reflected the dramatically lower growth rate of the active labour force, at 0.65 per cent per annum (against the national rate of 3.6 per cent). This contrast could have been even sharper for agriculture alone, since the period saw a sharp increase in the share of the rural labour force employed in non-agricultural rural activities; again, an aspect of income diversification reflecting the survival algorithms of rural households.

Agricultural output rose by 1.55 per cent per year or nearly a percentage point higher than the growth rate of the active rural labour force, which again contrasts sharply with the rest of the economy. The missing jigsaw piece is provided, of course, by heavy rural out-migration, partly into rural frontier areas, but substantially and increasingly into the urban sector, dominated by Manila. The differences in the trends in the real wage rates of unskilled workers in Manila and of farm workers is in fair measure due to this factor,5 though some role must be ascribed to the pattern of technological and institutional change in the countryside.

5 Naturally, the higher rates for the farm workers did not mean that the static Manila rate was unattractive in absolute terms.

Additionally, there are indications of diminishing returns in the process of agricultural intensification.6 This, coupled with declining per capita availability of land for farm operators partially explains the declining share of the rural workforce in agriculture. Between 1980 and 1991, Census data show a decline in the average farm area per farm operator from 2.8 hectares to 1.4 hectares (Sanchez, 1994: Table 2.10).

6 This is evident from sharply declining incremental gains in physical yields to fertilizer applications in palay production (Sanchez, 1994:Table 2.8). For fisheries, this is reflected in similar falls in the catch per unit of effort (in metric tonnes per trawler HP) (Cruz and Repetto, 1992:24, Table 2-10).

The sorry tale of the diverse forms of environment degradation which ran well into this period is well known and scarcely needs recounting. With the exception of Jamaica, the Philippines has the dubious distinction of having posted the highest annual rate of loss of natural forest cover during the 1980s, at 2.9 per cent (World Bank, 1994: 226-7, Table 33). The unsustainability of ecological and social unsustainability of agricultural intensification strategies in turn triggered off the heavy out-migrations. These took various forms. When intra-rural, they had frontier lands or uplands as their destinations; in the case of the latter, this meant only an extension of the process of degradation and migration. But increasingly, they came to be urban bound.7

7 This aspect is discussed at greater length in a subsequent section, where its links to the overseas migration phenomenon are explored.

Underlying this entire episode are two sets of fundamental causal factors, viz., first: anti-agricultural bias implicit in macro and sectoral policies; and second: the deep structural disarticulation of rural production and distributional systems. These need only be mentioned here, though it is difficult to over-emphasize their importance at an explanatory level. At the policy level, the anti-agricultural bias was manifest in the exchange rate and implicit tax regime designed to extract agricultural surpluses in favour of the protected industrial sector.8 Additionally, the share of agriculture languished and declined from a low base in the decade leading up to the reference period. Cruz and Repetto (1992: 27, Table 2-12) provide an estimate of natural resource depreciation, which when deducted from the gross investment from gross domestic investment in agriculture yields a substantial net disinvestment over the 1970-83 period. This clearly shows the weakening resource base of the rural sector under the influence of governmental policies.

8 On one calculation, the net resource transfers out of agriculture amounted to an annual average of 25 per cent of the agricultural value added during the period 1970-82 (see Bautista R.,1993:223, citing a study by Intal and Power).

Another aspect of the undervaluation of capital was the heavily labour displacing mechanisation in agriculture. Alongside this were institutional changes in traditional labour and employment systems involving harvesting and weeding operations which also involved a displacement of “ordinary” rural labour in favour of fewer, more fully employed operatives who could be more easily supervised. Such operatives were paid higher wages, and this could also partly explain the higher growth rates of real wages for farm workers for this period of stagnation.

On the structural side, the extremely high degree of landlessness, at approximately 50 per cent, needs recording. Juxtaposed on this is a second layer of sharp inequality in the distribution of total cultivated farm land. The Agrarian Reform Programme of 1987-97, had fulfilled only a very minor proportion of its targets thus far.9

9 During 1992, the achievements with respect to land area and the number of beneficiaries covered by the programme constituted 2.6 per cent and 3.7 per cent respectively of the total targets for the 10 year period. David (1992:11, Table 2) shows that up to 1986, the total achievement of land distribution targets had been just 3.2 per cent of the planned scope of the programmes; between 1987 and 1990, with respect to the Aquino initiated programmes, the cumulative achievements were no more than 13.8 per cent.

The second structural factor relates to agro-processing industry, with its massive bias in favour of large-scale, capital-intensive, mechanized, export oriented, multi-national plants. These reflect and farther accentuate the powerful biases of the price and trade regime mentioned earlier, and effectively exclude the peasant sector both from the rural industrial policies, and simultaneously from the employment which such plants could generate had they been more labour intensive. Furthermore, these industries display extremely weak linkages with the rural sector apart from using the basic raw materials, and thus generate only limited developmental spread effects. Such size-structure and scale biases tend to be carried over into monopolistic market domination, which farther accentuates the misallocation of resources, and the farther exclusion of small-scale independent suppliers from lucrative markets in urban centres. The Philippines case provides a textbook case in this regard. Alternative development strategies designed for the country have often taken this as a key point of emphasis.10

10 This problem has been thoroughly and lucidly analysed in Ranis, Stewart and Reyes (1989), where a comparison is made with Taiwan (China) where structural and policy features combined to generate a very different profile for the important export-oriented agricultural processing industry. In the case of Taiwan (China), this formed one of the planks of early, egalitarian growth.

There are several points to elicit from this extensive review of the rural sector and the analysis of the supply-side push factors generating “emigration pressures.” The first is that the extended period of stagnation, while not fully expressed equivalently in un- or underemployment, has created a very depressed domestic labour market situation characterized also by low quality employment. The length of the period of stagnation has been such that the cumulative pressures created could have been very considerable. Second, the unsustainability of domestic agricultural exploitation processes has eventually led to heavy rural out-migration flows which have gravitated disproportionately to the National Capital Region. This in turn, has reinforced the squeeze on urban labour markets, with obvious consequences for the prospects, quality and terms of employment for the “resident” urban workforce itself confronted by stagnant work opportunities.

It is necessary to register a few pertinent observations with respect to some urban labour market aspects of special relevance to the issue of emigration pressures. First, there is a clear inverse relationship between the rates of underemployment and unemployment across regions, and this is probably related to the degree of urbanisation; urban centres seem to record high unemployment and lower underemployment rates. This is true of the NCR where the rate of unemployment climbed in the first half of the period, before falling somewhat.11 Second, this pool of unemployed included a significant section of educated young persons. In 1985, a bad year, those with at least some college level education formed a staggering 38.5 per cent of the unemployed, bearing simultaneous testimony to the failure of the economy and the success of the educational system. In 1990, the share had dropped, but remained over the 30 per cent mark (Orbeta. Jr. and Sanchez, 1995: Table 16). If an epicentre of high emigration pressure had to be located for that period in the Philippines, it would surely have been located in this vast pool of qualified, young, aspiring but frustrated section of urban Philippines.

11 The rate rose by no less than 10 percentage points between 1982 and 1986; but then inexplicably dropped from 28.6 per cent to 18.3 per cent in just one year. This odd movement could have more than a little to do which the adoption of a new population series as from 1987. As such, both the sharply rising trend up to 1986, as well as the gentler decline from 1987 might retain statistical validity, though a comparison across 1986/87 would not.

The overall conclusion with regard to the supply-side factors must be that there is little likelihood that pressures emanating from this side of the equation will abate in the foreseeable future even in the face of moderately satisfactory macro-economic performance over the present plan period.