Documento(s) 11 de 29

Introduction

Textile manufacturing

The technology used in the mills is simple. Raw cotton, often compressed in bales, is mixed and blended, goes through various processes, (spinning, blowing, carding, drawing, roving, etc.), and emerges as yarn. The yarn is transmitted to the loom shed (weaving shed), where further processing produces grey cloth (unprocessed cloth). The grey cloth then goes to the processing shed for desizing, bleaching, dyeing, printing, and so on. The processed cloth is finally cut into suitable sizes and packed for sale. A crew of line operatives and helpers, supported by maintenance personnel and supervisory staff, forms the core of the labour in each of these steps.

The analysis in this paper is based on five spinning sheds and four weaving sheds. All the spinning sheds produce a uniform product (coarse yarns, mostly of count 20s), and all the weaving sheds produce a uniform product (standard plain weaves). These plants were established in the mid-1960s, and all are publicly owned.

Performance measures

Labour productivity

Capacity utilization

The situation was different in weaving: capital stock in Kiltex-Dar and Kiltex-Arusha remained constant during the 1973–79 period, but additional looms were installed in Urafiki sheds 1 and 2. The technical characteristics of these additional looms were identical to those of the original equipment, so the number of new looms was simply added to the number of the old looms to indicate the size of the capital stock. Change in the ratio of loom–hours to looms installed was, therefore, used to measure change in capacity utilization in the weaving sheds.

Machine productivity

Changes in output and labour productivity

Spinning

Table 1. Change in output in the spinning sheds (1973–79).
	Output (1000 kg)		Annual change (%)
	1973	1979
Urafiki-1 Urafiki-2 Mwatex Kiltex-Dar Kiltex-Arusha Group	3 218.0 1 651.0 2 891.4 1 728.1 897.2 10 385.7	3 693.6 1 933.2 2 778.7 1 128.3 860.0 10 393.8	2.11 2.44 –0.56 –4.96 –0.59 0.21
Source: The Production Statistics departments of the mills studied.

Table 2 shows (1) the level of person–hour inputs (actual and potential) in 1973 and 1979; (2) the levels of labour productivity for both years; and (3) the average annual rate of change in person–hour inputs and in labour productivity. The table shows that actual person–hour inputs for the group increased by 6.4% annually. The potential person–hour inputs increased at a faster rate during the period. The change in labour productivity (based on actual person–hour inputs) was marked — productivity fell in all the sheds by about 4–9% annually. The rate of productivity decline was higher on the basis of potential person–hour inputs.

Table 2. Change in actual and potential person–hour inputs and labour productivity in the spinning sheds (1973–79).
	Person–hours (thousands)		Annual change (%)	Productivity (kg/person–hour)		Annual change (%)
	1973	1979		1973	1979
Actual person–hour inputs and labour productivity
Urafiki-1 Urafiki-2 Mwatex Kiltex-Dar Kiltex Arusha Group	1952.1 930.5 1488.9 1006.9 548.9 5927.4	2817.3 1509.7 2528.4 909.7 824.4 8589.5	6.3 8.4 9.2 –1.7 7.0 6.4	1.65 1.77 1.94 1.72 1.63 1.74	1.31 1.28 1.10 1.24 1.04 1.19	–3.7 –5.3 –9.0 –5.3 –7.2 –6.0
Potential person–hour inputs and labour productivity
Urafiki-1 Urafiki-2 Mwatex Kiltex-Dar Kiltex-Arusha Group	2054.3 989.8 1628.3 1097.2 600.1 6369.7	3299.3 1728.1 2857.0 1051.9 914.4 9850.7	8.2 9.7 9.8 –0.7 7.3 7.5	1.70 1.67 1.78 1.07 1.50 1.63	1.12 1.12 0.97 1.07 0.94 1.05	–5.5 –6.4 –9.6 –6.3 –7.5 –7.1
Source: The Production Statistics departments of the mills studied. Note: PH, person–hours.

Weaving

Table 3. Change in grey-cloth output in the weaving sheds (1973–79).
	Output (1000 m)		Annual change (%)
	1973	1979
Urafiki-1 Urafiki-2 Kiltex-Dar Kiltex-Arusha Group	13 767.1 7 748.3 10 533.2 4 394.2 36 442.8	16 354.4 11 456.3 6 132.4 7 264.0 41 207.1	2.9 6.7 –8.6 8.7 2.1
Source: The Production Statistics departments of the mills.

Labour-productivity measures were based on both actual and potential person–hour inputs. Table 4 shows (1) the levels of these inputs in 1973 and 1979; (2) the levels of labour productivity in both years; and (3) the average annual rate of change in person–hour inputs and in labour productivity.

Actual person–hour inputs for the group as a whole increased at an average rate of about 9% annually. Among the individual sheds, the rates of increase varied widely. The rates of increase in person–hour inputs were roughly associated with rates of growth in grey-cloth output, but in all cases the growth in person–hour inputs was much greater. Consequently, the level of labour productivity for the individual sheds fell (albeit at different rates) during the 7-year period.

Table 4. Change in actual and potential person–hour inputs and labour productivity in the weaving sheds (1973–79).
	Person–hours (thousands)		Annual change (%)	Productivity (m/person–hour)		Annual change (%)
	1973	1979		1973	1979
Actual person–hour inputs and labour productivity
Urafiki-1 Urafiki-2 Kiltex-Dar Kiltex-Arusha Group	6 468.1 3 842.7 4 271.8 1 844.3 16 426.9	8 972.3 7 440.2 5 216.5 6 059.3 27 688.3	5.6 11.6 3.4 21.9 9.1	2.13 2.02 2.47 2.38 2.20	1.13 1.54 1.18 1.20 1.49	–2.6 –4.4 –11.6 –10.8 –6.4
Potential person–hour inputs and labour productivity
Urafiki-1 Urafiki-2 Kiltex-Dar Kiltex-Arusha Group	6 885.4 4 070.3 4 271.8 1 961.3 17 474.1	10 315.0 8 962.2 6 082.6 6 723.4 32 083.2	7.0 14.1 4.9 22.8 10.7	2.00 1.90 2.31 2.24 2.09	1.59 1.28 1.01 1.08 1.28	–3.8 –6.4 –12.9 –11.5 –7.0
Source: The Production Statistics departments of the mills studied. Note: PH, person–hours.

Changes in capacity utilization and machine efficiency

Spinning

Table 5. Change in spindle–hours worked in the spinning sheds (1973–79).
	Spindle–hours worked (n)		Annual change in spindle–hours (%)	Annual change in output (%)
	1973	1979
Urafiki-1 Urafiki-1 Mwatex Kiltex-Dar Kiltex-Arusha Group	265.47 125.44 235.85 140.23 76.42 843.41	374.90 186.52 250.37 115.65 79.56 1007.00	5.9 6.8 1.0 –3.2 0.7 3.0	2.11 2.44 –0.56 –4.96 –0.59 0.01
Source: The Production Statistics departments of the mills studied.

As might be expected, differences in the rate of change in capacity utilization among individual mills were associated with differences in the rate of change in output. For the group as a whole, spindle–hours increased at a much faster rate (3.0% annually) than output (0.01% annually). Clearly, then, output per spindle–hour was not rising; in fact, it was falling (Table 6).

Table 6. Change in spindle productivity in the spinning sheds (1973–79)
	Productivity (kg spindle–hour 1)		Annual change (%)
	1973	1979
Urafiki-1 Urafiki-2 Mwatex Kiltex-Dar Kiltex-Arusha Group	0.0121 0.0121 0.0123 0.0123 0.0117 0.0123	0.0098 0.0104 0.0111 0.0097 0.0108 0.0103	–3.5 –2.5 –1.6 –3.9 –1.4 –2.9
Source: The Production Statistics departments of the mills studied.

Evidently, although management increased the person–hour inputs in order to expand capacity utilization, this move did not raise the productivity of the running machinery. Indeed, machine productivity was not even held constant as capacity utilization increased. Because output during the period of study was more or less constant for the group as a whole, increasing capacity utilization (expanding by about 3.0% annually) was required simply to compensate for decreasing machine efficiency (falling by about 2.9% annually).

Weaving

Table 7. Change in the ratio of loom–hours operated to number of looms installed in the loom sheds (1973–79).
	Loom–hours		Change in ratio of loom–hours to looms (%)	Annual change in grey-cloth production (%)
	1973	1979
Urafiki-1 Urafiki-2 Kiltex-Dar Kiltex-Arusha Group	14 922 11 279 12 065 8 241 46 507	18 675 13 976 13 986 26 257 17 203	3.8 3.6 2.5 21.3 6.1	2.9 6.7 –8.6 8.7 2.1
Source: The Production Statistics departments of the mills studied.

In all the weaving sheds, individually and taken as a group, capacity utilization increased. Differences in the rate of change among individual mills were loosely associated with differences in the rate of change in output. Although capacity utilization was rising, output per loom–hour was not; in fact, it was rapidly decreasing (Table 8).

Table 8. Change in loom–hour productivity in the weaving sheds (1973–79).
	Productivity (m/loom–hour)		Annual change (%)
	1973	1979
Urafiki-1 Urafiki-2 Kiltex-Dar Kiltex-Arusha Group	2.05 2.08 2.09 2.04 2.07	1.38 1.42 1.05 1.06 1.28	–6.4 –6.2 –10.8 –10.3 –7.7
Source: The Production Statistics departments of the mills studied.

Evidently, then, although some of the increase in capacity utilization resulted in an increase in output, a much larger proportion was simply required to offset rapidly falling loom–hour productivity.

Benchmark efficiency levels

This path of development in one of the country's leading manufacturing sectors should perhaps be set in context. This analysis does not cover the initial start-up or running-in phase of production in the mills. By 1973, all the mills studied had been operating for at least 5 years, so the decline in productivity does not reflect a decline from design-level efficiencies that had been attained in the start-up phase.

It was possible to establish benchmark efficiency levels for the types of equipment installed in the mills. With the assistance of the staff of the Shirley Institute, based in Manchester in the United Kingdom, I estimated benchmark efficiency levels for two performance indicators (labour productivity and machine– hour productivity) as being 65% of the specified design-level efficiencies for the type of equipment used in Tanzania. The downward adjustment of 35% from the specified design-level efficiencies allows for start-up discrepancies and provides a reasonable norm for Tanzania.

Spinning

• labour productivity = 4.55 kg/person–hour; and

• machine productivity = 0.035 kg/spindle–hour.

Table 9. Actual productivity levels achieved in the spinning sheds compared with the benchmark productivity levels (1973–79).
	Actual productivity levels		Ratio of actual to benchmark (%)
	Labour (kg/person–hour)	Spindles (kg/spindle–hour)	Labour–hour	Spindle–hour
1973
Urafiki-1 Urafiki-2 Mwatex Kiltex-Dar Kiltex-Arusha Group	1.65 1.77 1.94 1.72 1.63 1.75	0.0121 0.0121 0.0123 0.0123 0.0117 0.0123	36 36 43 38 36 38	35 35 35 35 35 35
1979
Urafiki-1 Urafiki-2 Mwatex Kiltex-Dar Kiltex-Arusha Group	1.31 1.28 1.10 1.24 1.04 1.21	0.0098 0.0104 0.0111 0.0097 0.0108 0.0103	29 28 24 27 23 27	28 30 32 28 31 29

Weaving

Table 10. Actual productivity levels achieved in the weaving sheds compared with the benchmark productivity levels (1973–79).
	Actual productivity levels		Benchmark productivity level		Ratio of actual to benchmark (%)
	Labour (m/person–hour)	Looms (m/loom–hour)	Labour (m/person–hour)	Looms (m/loom–hour)	Labour–hour	Loom–hour
1973
Urafiki-1 Urafiki-2 Kiltex-Dar Kiltex-Arusha	2.13 2.02 2.47 2.38	2.05 2.08 2.09 2.04	6.5 6.5 5.2 7.8	5.85 5.85 4.55 6.50	33 31 48 31	35 36 46 31
1979
Urafiki-1 Urafiki-2 Kiltex-Dar Kiltex-Arusha	1.13 1.54 1.18 1.20	1.38 1.42 1.05 1.06	6.5 6.5 5.2 7.8	5.85 5.85 4.55 6.50	28 24 23 15	24 25 23 16

In 1973, the actual efficiency levels in the weaving sheds were about a third of the benchmark levels (with the exception of Kiltex-Dar, where it was nearly half). By 1979, the process of unlearning had reduced relative efficiency to around 25% of the estimated benchmark levels, even in the case of Kiltex-Dar, and to only about 15% in the case of Kiltex-Arusha.

In effect, then, after some 10–15 years of cumulative production experience, the mills were producing as if they were still in the start-up or running-in phase of development. Production efficiencies were still far below design-level efficiencies. The cumulative production experience had not automatically generated the efficiency improvements needed to bring performance up to even the benchmark levels. In fact, performance was moving away from, not toward, those levels.

Evidently, increasing production experience and the passage of time were associated not with improving production efficiency, but with decreasing production efficiency.

Conclusions

The main finding was that from 1973 to 1979, productivity (x efficiency) in this industry, far from improving, actually declined. Labour productivity and machine productivity, two of the performance measures used to indicate efficiency levels and trends, showed a persistent decline. Capacity utilization, on the other hand, increased in almost every mill.

This suggests a general deterioration in efficiency in use of the imported technology. Clearly, then, this industry shows no evidence of technological learning in the sense of endogenous execution and management of incremental technological changes or productivity improvement. Instead, the industry appears to have rapidly unlearned.

The above conclusions suggest that very limited assimilation, absorption, and mastery of imported technology took place in this sector of the importing economy during this period. It also suggests that there wasn't much effort in Tanzania to build up the technological and managerial skills, expertise, and related capabilities needed to improve productivity and efficiency in the industry.

Recommendations for further research and analysis

1. Our knowledge about how technological change bears on the process of industrial development of Tanzania is very limited.

2. There are few empirical data on Tanzanian realities to inform future policies, plans, strategies, and management of technological change and industrial development.

Systematic and in-depth studies

Studies on the determinants of technological change

Comparative studies

Studies linking technological change and technology transfer

Studies on technological change and productivity improvement should address linkages between technology transfer and technological change. Realistically, in technologically underdeveloped economies, technology transfer and technological change form a continuum, rather than a series of discrete, unrelated processes. It is important, therefore, that studies of technological change in such countries take into account this rather obvious point.

Reference

• Mlawa, H. 1983. The acquisition of technology, technological capability and technical change: A study of the textile industry in Tanzania. Science Policy Research Unit – Institute of Development Studies, University of Sussex, Sussex, UK. DPhil thesis.

Documento(s) 11 de 29