Defining infrastructure: a broad concept

1. Should policymakers take account of demographic factors in considering investments in infrastructure?Peter S. Heller (SAIS)Presentation to POPNET Conference on Population, Reproductive Health, and Economic DevelopmentDublin, January 16, 2009

2. Defining infrastructure: a broad concept Typically, an investment in basic structures (not machinery or equipment) • Spatially universal infrastructure: housing, water, sanitation, social services • Economically productive infrastructure: energy, ICT, irrigation, ports, transport (roads, railroads) • Spatially connective infrastructure within a country • Regional infrastructure (whether to regional markets or to global markets)

3. Distinguish New investment vs Rehabilitation investment for existing infrastructure vs. Operations and maintenance (O&M) • There are substitution possibilities: an increase in the quality of infrastructure investment changes level of demand and periodicity of demand for O&M, with obvious fiscal consequences

4. Four Issues to Examine • How might demographic factors influence the demand for infrastructure? • Have demographic factors played a key role in the past in influencing infrastructure investments? • What do future demographic trends suggest about infrastructure needs, particularly in LICs? • What might be the policy implications?

5. Issue 1 How might demographic factors influence the demand for infrastructure?

6. In principle, many demographic variables influence the need for infrastructure • Population size of a country or urban agglomeration • Population dynamics of a country: what stage of the demographic transition? Implies age structure • Age structure of the population of country or city • Share of young? Elderly? • Density: rural; smaller cities (<100,000; 100,000--1 million, rural settlements) vs. mega-cities • Extent of migration: urban-rural; international Note: basic econometric models typically include: population size, density, urbanization rate

7. Population size and age structure • Population size influences the demand for spatially universal service infrastructure (though influenced by economies of scale): • Water and sanitation • Basic social services infrastructure (number of schools, health facilities • Age structure of population: • Young population: obvious bias toward demand for educational infrastructure • Working age group: public infrastructure to provide complementary inputs for private sector productivity and job creation • Elderly population: need an elderly-friendly infrastructure

8. Dynamics of population structure • Stage of demographic transition determines appropriate composition of infrastructure investment • Not yet in transition: need to scale up educational and health infrastructure as well as provide infrastructure for jobs of growing labor force • Youth dependency rate still high--conflicting challenge between meeting needs of youth and infrastructural requirements for growth (e.g., Kenya, Vision 2030) • As overall dependency rates decline and growing labor force: need infrastructure to foster job creation; slowing growth of youth population

9. In principle, demographic transition can allow for higher savings and investment rate • Higher working age group share and lower dependency rate: potential for increased savings rate and higher rate of capital formation (including of infrastructure), and economic growth rate. • Such investments vital to absorb growing labor force • As with Asia, higher growth in labor force might prove attractive for foreign investment. Also financing infrastructure • But note difference between Latin America and Asia in level of investment associated with lower dependency rate in 80’s &90s. Higher savings may not materialize • So, higher growth may not materialize commensurate with higher level of productive age work force

10. Contrast Asia and Latin America: A lower dependency rate not necessarily associated with higher investment rate

11. Dynamics of population structure (cont) • Later in demographic transition, should observe an increased elderly share & falling youth share (note: some youth bubbles do occur) and possibly declining population • Downsizing may be necessary (too many schools? Infrastructure system too large &uneconomic?) • Need infrastructure appropriate for elderly • Can expect this to happen in Eastern and Western Europe looking forward; also rural China in future • Particularly a challenge in smaller cities where there may be diseconomies of scale associated with limited client population served by overbuilt infrastructure

12. Urbanization: Mixed evidence for its implications for infrastructure • Economies of scale and benefits of higher density in reducing per capita costs for water, sanitation, power, transport infrastructure, and possibly social services. • But demand for higher quality and quantity of services increases overall costs per capita in urban areas for spatially universal services. This is particularly the case for larger cities. • Differential in infrastructure costs per capita with rural areas may be less in smaller cities. • But lack of financial resources may prevent necessary infrastructure investments from occurring • Note the significant disparities within urban areas between urban poor and other groups; slums;

13. Evidence; relative costs in urban and rural areas of water Two offsetting forces • Capital costs and salaries in urban areas are much higher • But lower population densities and longer distances can imply a higher cost of providing rural populations with access to water. • Millennium Project: assumed that rural capital costs for boreholes, rainwater collection, and dug wells are about 40 percent of urban cost • But household connections and public stand posts in rural areas are assumed to be twice as expensive as in dense urban areas!

14. Recognition that urbanization will require infrastructure is only first step • As noted, two critical types of infrastructure: • Universal services infrastructure--water, sanitation; Also need for housing as infrastructure • Economically productive infrastructure (related to job creation--ICT, transport links for export, electricity • But how to approach the creation of urban infrastructure? • Many possible choices--very different cost profile! • Upgrading slums--upgrading housing, retrofitting infrastructure for water supply, sanitation, transport and energy services • Requires strong focus on networked technologies--sewers, piped water and electricity grids, storm drainage; • Also, water storage • New urban developments as alternatives to formation of new slums • In Transport: large scale road and rail-based transport infrastructure vs bus-based mass transit approaches

15. With urbanization, there is potential for unbundling Example: Sewerage • For large urban areas, unbundling a service area into parallel independent service zones, each with its own sewerage network leads to lower average diameter and average depth for the entire city • Leads thus to lower capital costs, stretching funds; also leads to management of smaller areas of service--thus, easier to manage • Example: Bangkok: a megacity where unbundling of sewerage has been successfully applied

16. Again, note the cost differentials cited by Millennium Project • Slum Upgrading: Average investment per person ($) over 15 years (including. physical improvements to housing stock, basic physical infrastructure--water, sanitation, drainage, road paving, and electricity) • Upgrading slums:: $670 or about $42 per beneficiary per year • Providing alternatives to formation of new slums: $400 ($26 p.b.p.a) • Implication: new settlements can be build for one third to one half lower than the cost of comprehensive upgrading. Source: Millennium Project • Cost of water storage infrastructure: to reach South African standards, need roughly $15-$70 per capita per year!

17. Urbanization by design? • Some countries have tried to proactively influence where urbanization occurs by investments in infrastructure • Sometimes successful, e.g., Korea and China • But also note failed cases: often around EPZs (e.g., Kenya) • May result in costly and inefficient infrastructure provision • Most countries have less proactive policies: urbanization: governments imply respond to forces in growing economy: agglomeration economies: job opportunities that pull in labor and/or weak agricultural sector that pushes labor out

18. Other infrastructural consequences of urbanization • The need to feed an urban population creates pressure for • infrastructure in rural areas to support domestic agricultural production and to facilitate transport of commodities from the rural areas to cities • Need the capacity to both supply a city (railways, roads) • As city grows, existing infrastructure may become inadequate and may need to be upgraded or replaced • Creates opportunity for investments in new approaches for supplying water or sanitation • Need to reconsider urban transport infrastructure

19. Other infrastructural consequences of urbanization (cont) • Increased urbanization, particularly coastal cities, puts pressure on ground water • Forces consideration of alternative sources of water & requires better urban water demand management • With ground subsidence, infrastructure needed to limit risks posed from storms, wind damage, storm surges • Higher likelihood of residual damage from the occasional extreme storm event. • Coastal protection infrastructure will be a key element in reducing risk. Climate change will exacerbate these forces (as will be discussed later)

20. Migration will also shape demand for infrastructure • Will migration be to smaller cities? To mega cities? To peripheral exurbs? • Sources of internal migration and effects on infrastructure needs in exporting areas • Endogeneity factor: provision of infrastructure may induce in-migration! But not strong evidence of this • Is immigration likely from neighboring countries? Conversely, will out-migration dampen domestic demographic pressures and influence population size and age structure? • All questions for policy maker to consider

21. Technological factors also shape level and character of demand for infrastructure • Technological change can induce or create demand for new kinds of infrastructure or substitutions, particularly for big ticket items: energy, transport, ICT • Dedicated urban bus lanes vs auto-driven transport network in urban areas • ICT: role of land lines vs. cable optic fibre vs cell phone towers; do conventional technologies become outmoded? Cell phone technologies leapfrogging over demands for landlines • Energy sector: Is there a shift towards renewables? • Is there a potential for economies of scale for a given type of infrastructure? • For certain types of infrastructure (power, transport), demographic variables (e.g., population size) may be dominant consideration) relative to technological factors

22. Policy makers should consider choices among existing technologies • Avoid adopting technical design standards from industrial countries. • Consider adaptation to local circumstances (e.g.,Brazil and Australia) • Millennium Project examples • Pour-flush systems (the Sulabh program in India) vs. flush toilets: pour-flush systems reduce quantity of water demanded and quantity of wastewater produced; low-volume-flush toilets are water-saving devices • Consider potential for labor-intensive roads to create employment and minimize adverse environmental impact • Water supply: choose infrastructure that matches occurrence and sources of water, treatment needs; socioeconomic status of intended users; and location and size;

23. Example: choosing among alternative water supply technologies: what size of city? • Large cities:rely on surface water: Construct dams: • For large scale systems, institutions, and domestic and small scale agricultural uses: can also rely on ground water; Dig boreholes and tube wells • For village or community or household uses: large diameter wells: dug wells or mechanically dug: • For rural areas: rely on ground water & spring water: protected spring box;

24. Example: selected technological options for sustainable access to sanitation • On-site sanitation: (function of how much water usage, permeability of soil, depth of water table, density of housing) Excreta Disposal • Simple unventilated double pit toilet • Pour-flush toilet with twin soak-away pits • Pour-flush toilet plus septic tank with twin-pit soak-away pits Wastewater disposal • Separate twin-pit soak-away system for sewage disposal • Off-site sanitation: function of amount of water usage, soil permeability, housing density, depth of water table • Wastewater conveyance: simple sewer system • Primary treatment: sludge drying beds and Imhoff tank • Secondary treatment: trickling filters , sludge digesters, co-composting of sludge with garbage • Alternative treatment options: constructed wetlands, in-stream wetlands and waste-stabilization ponds Source: UN Millennium Project Task Force

25. Energy • Millennium Project illustrates case of Brazil, where significant effort to move away from biomass production of energy through use of across-the-board and targeted subsidies to the poor for natural gas including in rural areas. • Implies a shifting of infrastructure strategy from existing technology to an alternative • Infrastructure requirements? Actually, principal upfront cost is the purchase of LPG cylinder and stove

26. Level of development and growth imperatives • Infrastructure can facilitate/stimulate growth • Provide complementary inputs to private sector • Key policy issue: what infrastructure appropriate at different phases of development? • Role of FDI • But reverse causality may apply: growth and rising PCI will stimulate demand for infrastructure • Need to upgrade infrastructure to meet pressures and demands for energy; ICT; water; transport networks, etc • Demand for higher standards of infrastructure

27. Stage of development itself fosters demand • In LICs and MICs, major differential between infrastructure in urban and rural areas. Often, a major backlog owing to previous lack of investment! • In industrial countries, observe a convergence in quality of spatially universal infrastructure (WSS, medical, roads, education) per capita in urban and rural areas. • Also, what is defined as “spatially universal” evolves. In MDCs, electricity becomes necessity! • WEO (2008) notes rapid growth of demand for car ownership as PCI approaches a given threshold: intensifies demand for associated infrastructure: urban and interurban

28. Convergence of demand for spatially universal infrastructure

29. Another factor to consider: the MDGs! • MDG 7 relates to “ensuring environmental sustainability” • Target 7c: Reduce by half the proportion of people without sustainable access to safe drinking water and basic sanitation • 7.8 Proportion of population using an improved drinking water source • 7.9 Proportion of population using an improved sanitation facility • Arguably Target 7d: Achieve significant improvement in lives of at least 100 million slum dwellers, by 2020 • 7.10 Proportion of urban population living in slums • Simply to meet MDGs, particularly with rapidly growing urban populations, need to invest significant sums in infrastructure • Millennium Project: estimates it would cost roughly $293.5 billion over the period 2005-2020, or roughly $20 billion annually (of which $11 billion from ODA). Averages 70-80$ per capita to meet all MDGs

30. Millennium Project estimates of annual investment costs per capita to meet MDGs (as opposed to universal provision!) For Bangladesh, Cambodia, Ghana, Tanzania, and Uganda • WSS: $5 - $7 per capita annually: more generally, $2-$6, rising to $6-$12 • Improve lives of slum dwellers $2 - $4 • Energy: $11-$19; more generally $6-$20 rising to $18-$23 • Roads: $21 • Note; these are rough averages; lower in beginning years, higher later on. Countries of course differ Source: Millennium Project Overview: chapter 17

31. Cost of just meeting water and sanitation MDGs • Global finance costs range from $51 billion to $102 billion for water supply; and from $24 billion to $42 billion for sanitation for the period 2001-15; average $68 billion for water and $33 billion for sanitation: aver annually $6.7 billion • No absolute cost figure--would depend upon the technologies adopted and country-specific preferences and conditions. • Estimates are for a minimum package of services, in which low service levels (in terms of technologies and costs) were applied for rural populations and intermediate service levels were applied for urban populations. The vast majority of need was assumed to be in periurban areas and slums.

32. Another factor influencing demand for infrastructure in future: Climate Change • Future climate change may significantly exacerbate pressures from urbanization. • Undercut viability of some areas for settlement • Increased risk of flooding, major disruption of trade • Looking forward, may engender migration or resettlement, creating new demands for infrastructure • May influence the viability of existing infrastructure • Difficult to separate out the pressures from demographic change from the associated developments that can accompany such demographic change • viz., subsidence; socioeconomic development; etc

33. A further consideration: do demographic considerations have adequate voice? • An important political economy question in low income countries particularly, is whether • The fact of demographic needs for infrastructure are articulated and responded to by policy decision-makers. If decisions on demography are made on technocratic or political grounds, demographic-related needs may not be taken into account; • Also, in many societies, there may be various sociological factors that ignore the needs or demands of important segments of the population. • Women, for example, may be effectively supplying infrastructural services (water, energy) and their needs may not be adequately considered by the political forces that shape infrastructural investment decisions • Women may also want preference given to technologies for infrastructure for which they can have some control, rather than large infrastructure projects

34. Finally: fiscal constraints limit whether infrastructure investments respond to demographic needs ! Infrastructure: usually a public good—subject to economies of scale; externalities; spillover of benefits (and costs) • Externalities often large and capacity for excludability may be limited--- such infrastructure may not be viable on commercial terms for much of the population • Observe only limited private investment in dams, power, highways, reservoirs; water and sanitation • Some exceptions: some infrastructure, e.g., satellites, telecommunications, can be undertaken privately; private willingness to pay; significant private investment developing (e.g., East Africa)

35. Fiscal constraints (continued) • At all PCI levels, fiscal constraints dictate adequacy, quality, and magnitude of the infrastructureto be provided • These constraints are more binding for LICs with low tax ratios, limited capacity for debt absorption, and heavy reliance on external assistance • Even where public-private-partnerships (PPPs) are used, fiscal contingent liabilities engendered; • Public policy thus critical in influencing what infrastructure choices are made: where?; how much? what technology? what policy issues dominate?

36. Issue 2 Have demographic factors played a key role in the past in influencing infrastructure investments?

37. Raw facts suggest otherwise. Look at Africa’s infrastructure deficit relative to other low income countries (Foster, 2008, p.2)

38. Recent World Bank Africa Infrastructure Diagnostic studies reveal that • Power consumption in Africa only 10% that in developing world and falling (124 kwh per capita per year) • Since 1990, little change in share of population with access to land-line telephones, flush toilets or piped water. Little improvement in population share with access to electricity (from 22% to 28%) • Rapid urban growth “leaving infrastructure service providers severely stretched,…resulting gap [in water and sanitation] filled by lower cost alternatives such as boreholes and pit latrines” Source: Foster 2008

39. Issue 3 What do projected demographic trends suggest about infrastructure needs, particularly in LICs?

40. A few key demographic factors to consider • Growth of population • Projected urbanization rates • Projected growth in the size of school age and working age populations • Projected growth of elderly--in LICs and MICs; in industrial countries

41. Africa: Seeking to catch up Asia: moving to the next level Compare two recent estimates of infrastructure needs over next 5 -10 years in Asia (Yepes, 2008) and Africa (Foster (2008)

42. In considering linkage between demographic projections and infrastructure needs, policy makers should differentiate need for: • Spatially universal infrastructure: WSS and infrastructure for basic social services (education and health) • Influenced by overall pop. growth & age structure (# of youth) • Recognize “need” and “effective demand” are different • Economically productive infrastructure: complementing workforce in manufacturing and services & facilitating growth and employment in urban areas • Influenced more by urban growth & demographic transition • Examples: power, transport, ITC, ports • Spatially connective infrastructure: WDR2009 emphasizes need to facilitate growth in non-urban areas through transport connections: rural-urban links

43. What do demographic projections portend about the overall need for infrastructure? • First, start with absolute growth of population between 2005-2025: Africa 482 million Asia 841 million • Whether in absolute population increase, level per capita of infrastructure, and even in terms of capital complementing increased workers, the major focus: Asia through 2025 • But, after 2025, Africa will be focus for the major increase in infrastructure required for economic growth • Minimal increase in working age group in Asia after 2025 and large increase in the number of elderly

44. Let us define three categories of country 1. Early in demographic transition 2. Mid-demographic transition 3. Late demographic transition

45. First category of countries: those now experiencing fastest rates of population growth : still early in demographic transition • Niger, Burundi, Liberia, Guinea-Bissau, Uganda, Mali, DR Congo, Chad, Afghanistan, Timor Leste, Yemen, Nigeria, Ethiopia, Kenya, Tanzania. • For these,a need for spatially universal services infrastructure, particularly in 2005-2025 • UN Population projections (med variant) assume gradual reduction in fertility rate over time. • If this holds, countries will observe classic early phase of the demographic transition: • falling dependency rates, • rising shares of the population in the 15-59 working age group. • Higher fertility assumption would imply need for additional infrastructure for schooling and less domestic resource availability with higher dependency burden!

46. Category 1 countries • 2005-2025: Still substantial rapid growth in population: much still in rural areas, including many in working age groups • 2025-2050: with fertility reduction: population growth rate falling and urbanization accelerating • See that growth in urban population largely concentrated in 15-59 age group

47. These are also countries where a significant percentage of the increased population will be in urban areas, particularly during 2025-2050, as the demographic transition takes hold • In some countries, particularly in 2025-50 the increased urban population will be dominated by the working age group, • Suggests relatively greater importance in later period of economically productive infrastructure —power, telecommunication, transport—that facilitates increased investments for services and manufacturing, relative to “basic services” infrastructure. • But there will still be demand for such infrastructure in earlier period--rising L force!

48. Rapid urbanization dominated by working age group population

49. Despite urbanization, for many countries, also a growing number living in the rural areas needing basic infrastructure • Particularly for next 20 years, before urbanization process in these countries picks up stream • Rural infrastructure needed both to address dramatic existing deficiencies &to respond to rural pop. growth • For rapidly growing population countries—Niger, Burundi, Guinea-Bissau, Uganda, Ethiopia, Kenya and to lesser extent Tanzania, DRCongo, and Afghanistan, spatiallyconnective transport infrastructure will also be important to facilitate demand for agricultural production from rural areas--supplied to urban areas • Numbers may be even larger if one accepts that median population variant probably too optimistic

50. Countries in early phase of demographic transition: rapid urbanization but rural areas still important