## _wp1066

## Source details

**Canonical URL:** [_wp1066](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1066.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1066.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1066.pdf.json)

---

### Purpose and scope
- Aim: assess macroeconomic impacts of advances in information and communication technologies (ICTs) in low- and low-middle-income countries.
- Focus: macroeconomic impacts of ICT-related capital deepening driven by falling prices of ICT equipment.
- Coverage: (near-)complete coverage of low- and lower-middle-income countries; principal time frames: 2001-06 (cross-section averages) and 1990-2006 (time-series).

### Analytical framework and data
- Model:
  - Growth-accounting model with two capital types: non-ICT-related and ICT-related.
  - ICT technological advances identified as falling relative prices of ICT capital (p2).
  - Cobb-Douglas production with two sectors; aggregate output Y = Y1 + p2 Y2.
  - Steady-state and perturbation (time-path) analyses used.
- Key assumptions and calibrations:
  - Equal factor shares of capital types across sectors in simplified representation.
  - Overall capital share α assumed = 0.5 (based on Senhadji (2000) country estimates; between median 0.47 and mean 0.52).
  - Depreciation rates adopted:
    - IT equipment: 31.5 percent.
    - Communications equipment: 11 percent.
    - Non-ICT capital: 6 percent.
- Data construction:
  - Nominal GDP, real GDP growth, aggregate investment: IMF World Economic Outlook (IMF, 2008).
  - Population and population growth: United Nations Population Division (via WEO).
  - ICT-related investment: estimated from trade data (COMTRADE) with a 72 percent mark-up for costs and taxes where domestic spending data unavailable; adjustments for domestic production where relevant.
  - Price indices: U.S. NIPA and PPI series; communications equipment series modified using Doms (2005) with extrapolation for missing years.

### Steady-state findings (selected results, 2001–06)
- Aggregate (97 countries):
  - ICT-related capital deepening contributed 0.301 (Percentage Points) to economic growth in 2001-06.
  - Contribution split: IT Equipment = 0.162 (Percentage Points); Communications Equipment = 0.139 (Percentage Points).
  - Investment (Percent of GDP) = 22.1; IT Equipment (Annual Growth, Percent) = 0.74; Communications Equipment (Annual Growth, Percent) = 1.39; Population growth (Percent) = 1.9; Real GDP Growth (Percent) = 5.1.
  - Elasticities of output: IT Equipment = 0.006; Communications Equipment = 0.007.
- Memorandum by income group (2001-06):
  - LIC:
    - IT Equipment (Percent of GDP) = 0.6; Communications Equipment = 1.3.
    - Elasticities: IT = 0.005; COM = 0.007.
    - Contributions: IT = 0.131; COM = 0.128; ICT = 0.259 (Percentage Points).
  - LMC:
    - IT Equipment = 1.0; Communications Equipment = 1.5.
    - Elasticities: IT = 0.007; COM = 0.008.
    - Contributions: IT = 0.213; COM = 0.151; ICT = 0.364 (Percentage Points).
- Country examples (2001-06, contributions in Percentage Points):
  - Bangladesh: IT = 0.11; Communications = 0.09; ICT = 0.20.
  - Egypt: IT = 0.05; Communications = 0.08; ICT = 0.14.
  - Ethiopia: IT = 0.17; Communications = 0.14; ICT = 0.31.
  - India: IT = 0.19; Communications = 0.16; ICT = 0.35.
  - Indonesia: IT = 0.09; Communications = 0.15; ICT = 0.24.
  - Nigeria: IT = 0.07; Communications = 0.11; ICT = 0.18.
  - Pakistan: IT = 0.13; Communications = 0.12; ICT = 0.25.
  - Philippines: IT = 0.18; Communications = 0.10; ICT = 0.28.
  - Vietnam: IT = 0.12; Communications = 0.15; ICT = 0.28.
- Cross-country variation:
  - For the 10 countries in Table 1, ICT equipment contribution ranged between 0.14 and 0.35 (Percentage Points).
  - Variance in contribution from IT equipment is 60 percent higher than the variance from communications equipment across full set.

### Time-evolution and perturbation (1990–2006) findings
- Elasticities over time:
  - Average elasticity of output w.r.t. IT equipment grew from 0.0025 in 1990 to 0.006 in 2006.
  - Average elasticity of output w.r.t. communications equipment grew from 0.006 in 1990 to 0.009 in 2006.
  - R2 of linear trend regressions: IT elasticity = 0.85; communications elasticity = 0.47.
- Price-decline fluctuations (annual rates):
  - IT equipment: range 8.8 percent to 26.4 percent.
  - Communications equipment: range 5.6 percent to 10.7 percent.
- Impulse dynamics:
  - Immediate impact of a shock to relative prices accounts for only 50 percent of long-run effects.
  - Half of the remaining adjustment takes about 13 years (typical half-time ≈ 13 years).
- Estimated contributions (1991–2006, unweighted averages):
  - All countries: Total ICT Equipment = 0.19 (1991-1995), 0.25 (1996-2000), 0.26 (2001-2006).
  - LIC: Total ICT = 0.15, 0.20, 0.21 (same periods).
  - LMC: Total ICT = 0.22, 0.30, 0.31 (same periods).
- Periodal patterns:
  - Contributions rose from about 0.09 percent early in period to about 0.13 percent by the end for IT and communications separately (aggregate contribution rising accordingly).
  - 1996–2000: peak in capital deepening contribution from declining IT prices.

### ICT production and production-side contributions
- General result: production-side productivity gains in ICT-producing sectors have miniscule growth impacts in most countries covered.
- Notable exceptions and magnitudes:
  - Vietnam: production of IT equipment impact exceeds 0.1 (Percentage Points) in 2001-05.
  - China, Philippines, Thailand: production of IT equipment impact exceeds 0.5 (Percentage Points) in 2001-05.
  - Philippines, 2001-05: total contribution of production of ICT equipment to growth = 2.2 (Percentage Points); net exports of electronic microcircuits accounted for 1.5 (Percentage Points).
  - China: net exports of communications equipment rose to about 1.5 (Percent of GDP) by 2005; including plausible domestic sales, production could exceed 2 (Percent of GDP), consistent with growth contribution of 0.2 (Percentage Points) or more.
- Measurement approach:
  - Include net exports of electronic microcircuits (SITC 2 category 7764) to capture component-level production and distinguish assembly from component production.
  - Use U.S. price-based measures of productivity gains (PPI and NIPA series) as proxies for productivity improvements in output deflators.

### Comparison with higher-income country datasets (Global Insight, 2001–05)
- Global Insight coverage and limitations:
  - High coverage for high- and upper-middle-income countries; extends through 2005 only; does not identify communications spending.
- Representative statistics (Global Insight sample, 2001-05):
  - 28 high-income countries: GDP per capita = 26,929 (U.S. dollars); IT-investment (Percent) = 1.18; IT-investment (Percent of GDP) = 5.6; IT-investment (Percent of total investment) = 0.27.
  - 18 upper-middle-income: GDP per capita = 4,478; IT-investment (Percent) = 1.14; IT-investment (Percent of GDP) = 5.8; IT-investment (Percent of total investment) = 0.25.
  - 18 lower-middle-income: GDP per capita = 1,502; IT-investment (Percent) = 0.73; IT-investment (Percent of GDP) = 3.3; IT-investment (Percent of total investment) = 0.16.
  - 5 low-income: GDP per capita = 624; IT-investment (Percent) = 0.72; IT-investment (Percent of GDP) = 4.1; IT-investment (Percent of total investment) = 0.16.
- Interpretation:
  - Impact of IT-related capital deepening similar for high- and upper-middle-income countries (0.27 and 0.25 Percentage Points, respectively).
  - Growth impacts about 0.1 (Percentage Points) lower for lower-middle-income and the few low-income countries covered by Global Insight, reflecting IT-investment about 0.4 (Percent of GDP) lower.

### Data limitations, sample exclusions, and sources of uncertainty
- Key data challenges:
  - Inconsistencies in national accounts and lack of disaggregated national accounts data identifying ICT investment in low- and lower-middle-income countries.
  - Trade data noisiness (especially for small countries); mark-up conversion (72 percent) from net imports to spending introduces noise and potential cross-country variation.
  - Use of U.S. price indices may mis-measure relative prices if production composition differs materially.
  - Missing or implausible data led to exclusion of certain countries or replacement by sample averages (e.g., Afghanistan data relate to 2002-06; Paraguay excluded from totals due to severe inconsistencies; Kiribati and São Tomé & Príncipe treated as outliers).
- Sample adjustments and exclusions:
  - Countries dropped for cross-country analysis: Afghanistan, Liberia, Timor-Leste (early GDP data gaps); Lesotho, Namibia, Swaziland (trade aggregation issues); Kiribati, Paraguay, São Tomé & Príncipe (large outliers).
  - Countries excluded from main analysis due to data limitations: Democratic People's Republic of Korea, Liberia, Somalia, Federated States of Micronesia, Iraq, Marshall Islands, West Bank and Gaza, Timor-Leste.
- Caveats on interpretation:
  - Steady-state analysis can exaggerate immediate impacts and miss lags; perturbation analysis shows substantial sluggishness (half-life ≈ 13 years).
  - Elasticities estimated via steady-state approach may be influenced by temporary investment accelerations; perturbation calibration allows elasticities to trend over time.
  - High contribution of ICT-related production to GDP does not automatically translate to national income gains due to terms-of-trade effects from falling ICT prices.

### Key quantitative facts and parameters (preserved)
- Overall capital share α = 0.5 (assumed).
- Depreciation rates: IT equipment = 31.5 percent; Communications equipment = 11 percent; Non-ICT capital = 6 percent.
- Mark-up to convert net imports into spending = 72 percent.
- Average elasticity of output w.r.t. IT equipment: 0.0025 in 1990 → 0.006 in 2006.
- Average elasticity of output w.r.t. Communications equipment: 0.006 in 1990 → 0.009 in 2006.
- Annual price-index relative prices (1990=100) for computers and peripheral equipment (NIPA): 1990: 100.0; 1991: 86.7; 1992: 71.9; 1993: 59.7; 1994: 51.4; 1995: 42.1; 1996: 31.5; 1997: 23.9; 1998: 17.6; 1999: 13.4; 2000: 11.4; 2001: 9.2; 2002: 7.8; 2003: 6.8; 2004: 6.2; 2005: 5.3; 2006: 4.5.
- NIPA/PPI series for communications and microcircuits listed in Appendix Table 5 (1990=100) with exact annual values reported in source.
- Aggregate average contribution (97 countries, 2001-2006): ICT Equipment = 0.301 (Percentage Points) to annual GDP growth (IT = 0.162; Communications = 0.139).
- Long-run vs immediate impact: immediate impact ≈ 50 percent of long-run; half of remaining adjustment occurs over ≈ 13 years.
- Country-level production and trade highlights:
  - China net exports of communications equipment ≈ 1.5 (Percent of GDP) by 2005; domestic production plausibly > 2 (Percent of GDP).
  - Philippines, 2001-05: production contribution = 2.2 (Percentage Points); electronic microcircuits net exports = 1.5 (Percentage Points).

### Policy-relevant implications (implicit from analysis)
- Falling relative prices of ICT equipment have nontrivial but modest direct contributions to growth in low- and lower-middle-income countries (typical aggregate contributions ~0.2–0.3 percentage points over 1990–2006).
- Benefits unfold over long horizons due to capital accumulation lags (half-life ≈ 13 years), so short-run assessments understate long-run gains.
- Accurate measurement of ICT impacts requires:
  - Better disaggregation in national accounts to identify ICT production and investment.
  - National price indices that capture ICT quality improvements.
  - Careful treatment of trade-based estimates and mark-up conversions.
- Where domestic ICT production (especially components like microcircuits) exists, production-side productivity gains can materially raise measured growth contributions; measurement should distinguish assembly-intensive from component-producing activities.

*Source: IMF working paper — _wp1066 (selected sections and appendices) — author’s calculations and data as reported in the supplied document.*

### 1. Impact of ICT-Related Capital Deepening on Growth, Selected Countries, 2001-06 .........18

### 1. Impact of ICT-Related Capital Deepening on Growth, Selected Countries, 2001-06

### Purpose and scope
- Aim: provide an assessment of the macroeconomic impacts of advances in information and communication technologies (ICTs) in low- and low-middle-income countries.
- Focus: macroeconomic impacts of ICT-related capital deepening, i.e., implications of falling prices of ICT equipment for capital accumulation and economic growth.
- Coverage: designed to achieve (near-) complete coverage of low- and lower-middle-income countries.
- Time frames referenced in the study: 2001-06 (primary cross-section averages) and 1990-2006 (time-series analysis).

### Analytical framework and data constraints
- Framework: growth-accounting model with two types of capital — non-ICT-related and ICT-related — where technological advances in ICTs are identified as falling relative prices of ICT capital.
- Data inputs used:
  - Commonly available macroeconomic variables (e.g., investment rates, underlying growth rates).
  - International data on relative prices of ICT equipment.
  - Observed patterns of trade and, where available, production data to infer levels of ICT-related investment.
- Adaptation: framework adapted to the lack of disaggregated national accounts or industry data typical of advanced-country studies.
- Production data limitation: production data do not identify the role of ICT-related inputs, weakening the ability to assess the macroeconomic role of domestic ICT production in most low- and lower-middle-income countries.

### Channels of ICT impact considered and excluded
- Included channel: ICT-related capital deepening (absorption of ICT equipment such as computers and communication devices, and existence of telecommunications infrastructure).
- Complementary channel (assessed where relevant): contribution of ICT-producing sector to economic growth for the limited number of developing countries where production of ICT equipment is significant.
- Excluded channel (not quantified due to data constraints): generalized productivity gains associated with structural changes enabled by ICT usage. The paper notes that even for the U.S. the empirical evidence for broad-based productivity gains is weak and, given weaker statistical systems in the target countries, these effects are not estimated here.

### Empirical approach and analyses conducted
- Cross-section (2001-06): uses 2001-06 averages of key variables and draws on steady-state properties of the model to estimate contribution of ICT-related capital deepening to growth.
- Time-series (1990-2006): distinguishes immediate impacts of falling relative prices of ICT equipment on growth and “multiplier effects” that arise if a shock to growth affects subsequent rates of capital accumulation.
- Use of trade and production signals: in absence of disaggregated investment data, the study infers ICT-related investment from trade patterns and production where available.

### Key structural remarks from the introduction
- The macroeconomic impact of advances in ICTs is linked to the absorption of ICT equipment and telecommunications infrastructure.
- In most low- and lower-middle-income countries, production of ICT equipment does not play a macroeconomic role; in a few countries it is a significant contributor to growth.
- The analytical framework can be applied to estimate growth contributions from ICT production, but data weaknesses (lack of identification of ICT-related inputs) are a major obstacle.

*Source: IMF working paper — 1. Impact of ICT-Related Capital Deepening on Growth, Selected Countries, 2001-06 (introduction and overview material).*

### Section VI concludes. A data appendix offers a more extensive discussion of the construction

### _wp1066 - Section VI concludes. A data appendix offers a more extensive discussion of the construction

### II. Background
- Literature context:
  - Builds on the tradition established by Griliches and Jorgenson (1966, 1967).
  - Studies addressing ICT role in the U.S. growth resurgence in the latter half of the 1990s include Gordon (2000), Jorgenson (2001, 2005a), Jorgenson, Ho, and Stiroh (2005), Oliner and Sichel (2000), Oliner, Sichel, and Stiroh (2007).
  - Cross-country ICT impact studies: Jorgenson ((2003, 2005b)), Ahmad, Schreyer, and Wölfl (2004), Colecchia and Schreyer (2002), Pilat and Wölfl (2004), Daveri (2002), van Ark, O’Mahoney, and Timmer (2008), Bayoumi and Haacker (2002) (49 countries), Jorgenson and Vu (2005a, 2005b, and 2007) (110 countries).
- Key data challenges:
  - Inconsistencies in national accounts data across countries.
  - Lack of disaggregated national accounts data identifying production of or investment in ICT equipment in low- and lower-middle-income countries.
- Price index approach in prior literature:
  - Many studies construct “harmonized” price indices following Schreyer (2000, 2002), based on differences between prices of ICT equipment and non-ICT equipment in U.S. national accounts.
- Prior global datasets:
  - Jorgenson and Vu use sales data from the “World Information Technology and Services Alliance” (WITSA) available for 70 countries and extrapolate for another 40 using Penn World Table Version 6.1 (Heston, Summers, and Aten, 2002).
  - Jorgenson and Vu dataset covers only 50 of 103 countries classified as low- or low-middle-income by World Bank, 2007.
- Approach of the present study:
  - Construct a database of spending on ICT equipment based on trade data (with modifications for countries producing ICT equipment), using reported country data augmented by trade partner data where necessary.
  - Dataset covers all 103 low- and low-middle-income countries, going back to 1980 (coverage lower for earlier years due to country independence dates).
  - Focused analysis on contributions of production of ICT equipment and ICT-related capital deepening to economic growth rather than a full growth accounting decomposition.
  - Coverage of the growth impact analysis: 1990-2006, capturing between 89 and 97 countries towards the end of the sample period, and 80 countries at the beginning.
  - All spending on ICT equipment is attributed to investment in the present study.

### III. Analytical framework
- Objectives:
  - Develop a model to assess growth impacts of advances in ICTs.
  - Calibrate key parameters not directly observable from data.
- Focus areas:
  - Growth impacts of rising productivity in production of ICT equipment.
  - Contribution of ICT-related capital deepening fueled by declining relative prices of ICT equipment.
- Model structure:
  - Two-good framework: ICT products (subscript 2) and other products (subscript 1); good 1 is numeraire (p1 ≡ 1), p2 is relative price of ICT products.
  - Aggregate output: Y(·) = Y1(·) + p2 Y2(·).
  - Cobb-Douglas production functions with constant returns to scale:
    - Equation (2) specification: Y = A1 K11^α11 K12^α12 L1^(1-α11-α12) + p2 A2 K21^α21 K22^α22 L2^(1-α21-α22) (as presented in text).
  - Direct growth contribution of productivity gains in ICT sector: growth rate of A2 weighted by ICT sector share in output (Eq. (3)).
- Key assumptions to capture ICT-related capital deepening:
  1. Productivity gains in ICT production beyond non-ICT lead to equiproportionate decline in price of ICT equipment: x ≡ - (d p2 / dt) = growth rate of A2 minus growth rate of A1 (Eq. (4)).
  2. Factor shares of capital types 1 and 2 are the same across sectors.
- Simplified representation under equal factor shares:
  - Aggregate output in numeraire: Y = A p K1^α1 K2^α2 L^(1-α1-α2) (Eq. (5) and (6) per-capita y = A k1^α1 k2^α2).
- Capital accumulation (for capital type j ∈ {1,2}):
  - k̇j = sj y - (n + δj) kj adjusted for relative price p_j where relevant (Eq. (7)).
  - Integrated stock expression (Eq. (9)) and growth rate of capital stock expression (Eq. (10)).
- Role of relative price p2:
  - Faster decline in p2 increases growth rate of real ICT capital stock; identical nominal investment purchases more real ICT capital when p2 is lower.

### Steady-state properties of the model
- Steady-state definitions:
  - Constant sj, and constant growth rates γj of kj, y, and pi.
  - Eq. (12): γj = -δj + (sj y / kj) + g + n (as rearranged in text).
  - From steady-state, growth rates of capital stocks: k̇1/k1 = g and k̇2/k2 = g + x (Eq. (13)).
- Capital-output ratios in steady state:
  - k1/y = s1 / (g + n + δ1) (Eq. (14)).
  - k2/(p y) = s2 / (g + n + δ2 + x) (Eq. (15)).
- Elasticities and factor shares:
  - Overall capital share α = α1 + α2 (Eq. (16)).
  - No-arbitrage condition equating returns across assets yields relationship (Eq. (17) and (18)).
  - Solving yields expression for α2 in terms of observable parameters g, x, δ1, δ2, n, s1, s2 (Eq. (20)).
- Link between falling ICT prices and growth:
  - Aggregate impact on growth: dg/dx = α2 + (α1 + α2)(α2/(1-α2))? (represented in text by Eqs. (21) and (22) preserving exact formulation: dg/dx ≡ α2 + (α1/(1-α))? — text yields Eq. (21) and final relation Eq. (22): α2/(α1+α2) = dg/dx?).
  - Text conclusion: overall impact can be decomposed into direct impact via α2 and indirect impact via induced faster accumulation represented by α2(1-α1) terms; warns steady-state assumption can exaggerate immediate impacts and miss lagged impacts.

### Impact of ICT-related innovations over time
- Approach:
  - Treat falling prices of ICT equipment as time-variant disturbances γt to growth around a steady-state path (discrete-time perturbation approach).
  - Define γt such that γt = (2,2/2,1) /? (as presented in Eq. (23): γt ≡ (2,2/2,1)/? — preserve exact rendering: Eq. (23) given as 22,    2,1 /γα − = −∂ t t t pp where “∂” denotes deviation from steady-state).
- Aggregate capital-focused analysis:
  - Perturbation to output growth ∂y_t / y_{t-1} = γt + induced effects from capital accumulation (Eq. (24)).
  - Capital accumulation difference equation: k_{t+1}/k_t - 1 = s y_t / k_t - (n + δ) (discrete form Eq. (25)).
  - Under constant savings assumption, infinite-horizon expression (Eq. (26)).
- Induced perturbation dynamics:
  - Derived distributed-lag representations:
    - Induced perturbation to growth rate of capital as function of past disturbances γ (Eqs. (27)–(32)).
    - Change in output growth as immediate impact γt plus transitional impacts of past disturbances via capital accumulation (Eq. (33)).
- Long-term impact of a one-off disturbance γt:
  - Long-term impact on capital: sum of distributed lags (Eq. (34)).
  - Long-term impact on output: expressed in Eqs. (35) and equivalent expression (35) using price deviations; steady-state correspondence noted.

### IV. The contribution of ICT equipment to economic growth
- Analytical focus:
  - Evaluate growth impacts of ICT-related capital deepening in low- and lower-middle-income countries, given most do not produce ICT equipment.
  - Use steady-state framework for contributions in 2001-06 and time-evolution approach for 1990-2006 allowing elasticity to evolve.
- Data sources and construction:
  - Nominal GDP, real GDP growth, and aggregate investment from IMF’s World Economic Outlook database (IMF, 2008).
  - Population size and population growth from United Nations Population Division, included in World Economic Outlook database.
  - ICT-related investment estimated per data appendix:
    - For most countries, estimated from net imports of IT- and communications-related investment with a mark-up for costs and taxes not in net imports.
    - For some countries, based on available spending data.
    - For a few countries, adjusted for domestic production of ICT equipment.
  - Appendix Table 1 includes average investment rates for years 2001-05 (referred to; table content not reproduced here).
- Crucial variable:
  - The rate of decline of relative prices of ICT-related equipment is crucial for magnitude of growth effects (discussion continues beyond supplied excerpt).

*Italic source: _wp1066 - Section VI concludes. A data appendix offers a more extensive discussion of the construction*

### appendix for more discussion). The price series we use (Appendix Table 5) are based on

### Appendix: Price Series, Technological Parameters, and Steady-State Analysis (excerpts)

### Price series and data sources
- Price series are based on price indices from the U.S. National Income and Product Accounts and Producer Price Indices.
- For communications equipment, a modified series based on Doms (2005) is adopted; the Doms (2005) series do not cover the whole 1990-2005 period, and extrapolation is adopted for the outer years.
- Primary data sources referenced: IMF (2008), Global Insight (2006), UN Statistics Division (2008), U.S. Department of Commerce, Bureau of Economic Analysis (2008), and U.S. Department of Labor, Bureau of Labor Statistics (2008).

### Technological parameters used in the analysis
- Three technological parameters required:
  - overall factor share of capital α (= α1 + α2),
  - depreciation rate of non-ICT equipment δ1,
  - depreciation rate of ICT-equipment δ2.
- Choice of α:
  - Empirical studies commonly assume α in the range 0.35–0.40.
  - Senhadji (2000) provides estimates for 24 countries of interest; based on Senhadji (2000) the authors assume α = 0.5 (between the median 0.47 and the mean 0.52 of the country-level estimates).
- Depreciation rates adopted (following Jorgenson and Stiroh and Jorgenson, Ho and Stiroh):
  - IT equipment (“computers and peripheral equipment”): depreciation rate 31.5 percent.
  - Communications equipment (excluding structures related to telecommunications): depreciation rate 11 percent.
  - Non-ICT capital: depreciation rate 6 percent (reflecting inclusion of residential and other structures; Jorgenson and Stiroh propose depreciation mostly between 1 and 5 percent for structures).

### Steady-state analysis: setup and limitations
- Caveats about steady-state approach:
  - High rates of price decline and physical depreciation imply past investments have low weight in the stock of ICT capital, so averages may approximate stock of ICT capital.
  - Estimates may be misleading due to cyclical factors (business cycles, post-conflict recovery) causing unusual levels of ICT investment.
  - The role of ICTs may evolve over time; steady-state analysis would not capture evolving relations between ICT investment and growth.
- Scope of steady-state analysis:
  - Conducted for 97 countries with key macroeconomic data available over the 2000-2006 period (for a few countries overall investment rates were not available and sample averages substituted; Afghanistan relates to 2002-06).
  - Elasticities of output with respect to IT equipment and communications equipment are based on country-level estimates of GDP growth and population growth based on IMF (2008).
  - Analytical framework relies on Eqs. (20) and (22) (as referenced in source).

### Key steady-state findings (selected countries, 2001-06)
- Overall estimate:
  - ICT-related capital deepening contributed 0.3 percentage points to economic growth in 2001-06, with just over one-half attributed to IT equipment.
- Table 1 (selected country-level elasticities and contributions)
  - Bangladesh
    - IT Equipment (Percent of GDP): 0.52
    - Communications Equipment (Percent of GDP): 0.91
    - Elasticity of output w.r.t. IT Equipment: 0.004
    - Elasticity of output w.r.t. Communications Equipment: 0.005
    - Contribution to Growth — IT Equipment (Percentage Points): 0.11
    - Contribution to Growth — Communications Equipment (Percentage Points): 0.09
    - Contribution to Growth — ICT Equipment (Percentage Points): 0.20
  - China, P.R.: Mainland
    - IT Equipment (Percent of GDP): 0.46
    - Communications Equipment: ...
    - Elasticity of output w.r.t. IT Equipment: 0.003
    - Elasticity of output w.r.t. Communications Equipment: ...
    - Contribution to Growth — IT Equipment (Percentage Points): 0.10
    - Contribution to Growth — Communications Equipment: ...
    - Contribution to Growth — ICT Equipment: ...
    - (For China, data on investment in communications equipment were unavailable.)
  - Egypt
    - IT Equipment: 0.23
    - Communications Equipment: 0.80
    - Elasticity IT: 0.002
    - Elasticity COM: 0.004
    - Contribution IT: 0.05
    - Contribution COM: 0.08
    - Contribution ICT: 0.14
  - Ethiopia
    - IT Equipment: 0.80
    - Communications Equipment: 1.52
    - Elasticity IT: 0.006
    - Elasticity COM: 0.007
    - Contribution IT: 0.17
    - Contribution COM: 0.14
    - Contribution ICT: 0.31
  - India
    - IT Equipment: 0.88
    - Communications Equipment: 1.76
    - Elasticity IT: 0.006
    - Elasticity COM: 0.009
    - Contribution IT: 0.19
    - Contribution COM: 0.16
    - Contribution ICT: 0.35
  - Indonesia
    - IT Equipment: 0.41
    - Communications Equipment: 1.48
    - Elasticity IT: 0.003
    - Elasticity COM: 0.008
    - Contribution IT: 0.09
    - Contribution COM: 0.15
    - Contribution ICT: 0.24
  - Nigeria
    - IT Equipment: 0.36
    - Communications Equipment: 1.34
    - Elasticity IT: 0.002
    - Elasticity COM: 0.006
    - Contribution IT: 0.07
    - Contribution COM: 0.11
    - Contribution ICT: 0.18
  - Pakistan
    - IT Equipment: 0.58
    - Communications Equipment: 1.20
    - Elasticity IT: 0.004
    - Elasticity COM: 0.006
    - Contribution IT: 0.13
    - Contribution COM: 0.12
    - Contribution ICT: 0.25
  - Philippines
    - IT Equipment: 0.84
    - Communications Equipment: 0.96
    - Elasticity IT: 0.006
    - Elasticity COM: 0.005
    - Contribution IT: 0.18
    - Contribution COM: 0.10
    - Contribution ICT: 0.28
  - Vietnam
    - IT Equipment: 0.58
    - Communications Equipment: 1.67
    - Elasticity IT: 0.004
    - Elasticity COM: 0.008
    - Contribution IT: 0.12
    - Contribution COM: 0.15
    - Contribution ICT: 0.28
- Memorandum items (totals relate to all countries covered by analysis; totals exclude China and Paraguay)
  - All countries
    - IT Equipment (Percent of GDP): 0.74
    - Communications Equipment (Percent of GDP): 1.39
    - Elasticity IT: 0.006
    - Elasticity COM: 0.007
    - Contribution IT: 0.16
    - Contribution COM: 0.14
    - Contribution ICT: 0.30
  - LIC
    - IT Equipment: 0.6
    - Communications Equipment: 1.3
    - Elasticity IT: 0.005
    - Elasticity COM: 0.007
    - Contribution IT: 0.13
    - Contribution COM: 0.13
    - Contribution ICT: 0.26
  - LMC
    - IT Equipment: 1.0
    - Communications Equipment: 1.5
    - Elasticity IT: 0.007
    - Elasticity COM: 0.008
    - Contribution IT: 0.19
    - Contribution COM: 0.16
    - Contribution ICT: 0.35

- Additional steady-state insights:
  - For the 10 countries in Table 1, the contribution of ICT equipment to growth ranges between 0.14 and 0.35 percentage points (Egypt and India, respectively).
  - Variations in IT-related investment explain the bulk of cross-country variation in ICT contribution to growth:
    - Contribution from IT investment ranges between 0.05 and 0.19 percentage points.
    - Contribution from communications equipment ranges from 0.08 to 0.16 percentage points.
    - Across the full set of countries, the variance in the contribution of IT equipment to growth is 60 percent higher than the variance in the contribution of communications equipment.

### Comparison with advanced economies (Global Insight dataset, 2001-05)
- Methodological note:
  - Global Insight’s Global IT Navigator has high coverage of high- and upper-middle-income countries, extends through 2005 only, and does not identify spending on communications equipment; analysis focuses on IT-related capital deepening and period 2001-05.
- Table 2 summary (Impact of IT-Related Capital Deepening on Growth, Global Insight dataset, 2001-05)
  - 28 high-income countries
    - GDP per capita (U.S. dollars): 26,929
    - IT-investment: (Percent): 1.18
    - IT-investment (Percent of GDP): 5.6
    - IT-investment (Percent of total investment): 0.27
  - 18 upper-middle-income countries
    - GDP per capita (U.S. dollars): 4,478
    - IT-investment: (Percent): 1.14
    - IT-investment (Percent of GDP): 5.8
    - IT-investment (Percent of total investment): 0.25
  - 18 lower-middle-income countries
    - GDP per capita (U.S. dollars): 1,502
    - IT-investment: (Percent): 0.73
    - IT-investment (Percent of GDP): 3.3
    - IT-investment (Percent of total investment): 0.16
  - 5 low-income countries
    - GDP per capita (U.S. dollars): 624
    - IT-investment: (Percent): 0.72
    - IT-investment (Percent of GDP): 4.1
    - IT-investment (Percent of total investment): 0.16
  - Memorandum items:
    - 50 lower-middle-income countries
      - GDP per capita (U.S. dollars): 1,748
      - IT-investment: (Percent): 0.88
      - IT-investment (Percent of GDP): 3.8
      - IT-investment (Percent of total investment): 0.20
    - 47 low-income countries
      - GDP per capita (U.S. dollars): 409
      - IT-investment: (Percent): 0.59
      - IT-investment (Percent of GDP): 3.0
      - IT-investment (Percent of total investment): 0.13
- Interpretation:
  - Impact of IT-related capital deepening is similar between high- and upper-middle-income countries (0.27 and 0.25 percentage points, respectively).
  - Growth impacts are about 0.1 percentage points lower for lower-middle-income countries and the few low-income countries covered by Global Insight, reflecting IT-related investment is 0.4 percent of GDP lower in these countries compared to high- and upper-middle-income countries.
  - The authors find that Global IT Navigator may yield misleading results for low- and lower-middle-income countries due to limited coverage; adapting authors’ estimates to 2001-05 suggests growth impacts in low-income countries are about half of those in high- and upper-middle-income countries, with lower-middle-income countries about half-way between.

### Cross-country variation and time evolution
- Figure 1 summary (country-level variation):
  - Substantial variation in growth impacts of ICT-related capital deepening across low- and lower-middle-income countries with similar GDP per capita.
  - For IT-related capital deepening, growth impact appears correlated with GDP per capita; low-income countries cluster at growth impacts less than 0.2 percentage points.
  - For communications equipment, the cross-country pattern is less clear.
  - Note: Figures truncated for comparability; some country estimates (Kiribati and Paraguay) exceed truncation limits and are not shown in the figures; Appendix Table 1 contains full details.
- Contribution of ICT-related capital deepening to growth, 1990-2006 — limitations of steady-state approach:
  - The role of ICT equipment in the economy may change over time; an acceleration in ICT investment since 1990 implies an increase in the elasticity of output w.r.t. ICT equipment.
  - Pace of innovation (rates of price decline) is not constant; perturbations around steady-state can have long time lags, so steady-state may exaggerate immediate impacts and miss lagged impacts.

### Data issues, sample exclusions, and adjustments
- Countries dropped due to data limitations:
  - Afghanistan, Liberia, Timor-Leste (GDP data available only from 2000 or 2002).
  - Lesotho, Namibia, Swaziland (UN Comtrade subsumes trade data under Southern African Customs Union prior to 2000).
  - Kiribati, Paraguay, São Tomé & Príncipe dropped from cross-country analysis as ICT-related equipment series had large outliers.
    - Paraguay: IT equipment estimates (based on net imports) rise from 2.8 percent of GDP to 17 percent of GDP by 2006; published trade data show substantial inconsistencies.
    - Kiribati (1992) and São Tomé & Príncipe (1994) show isolated-year communications investment near 30 percent (treated as outliers).
- Missing investment rates:
  - Where investment rates unavailable, unweighted sample averages for investment rates for respective year were substituted (sample averages exclude Haiti and Macedonia, where IMF WEO investment rates exceed 100 percent of GDP for some years).
- Missing early-1990s macro data for some former Yugoslav and Soviet republics handled via simple extrapolations.

### Evolving elasticities of output with respect to ICT capital (calibration approach)
- Rationale:
  - Elasticities of output with respect to IT and communications equipment are allowed to vary across countries and over time to capture evolving role of ICTs.
  - Investments in both IT-related and communications equipment roughly doubled as a percentage of GDP between 1990 and 2006.
  - Notable differences: IT-related investment accelerated markedly in lower-middle-income countries relative to low-income countries; communications equipment investment rose markedly in low-income countries since 2002, catching up by 2006.
- Calibration steps:
  1. Calculate elasticities for each country and period using steady-state approximations (Eq. 20).
  2. Obtain linear trends through regressions on sample averages for each period of elasticities computed in step 1.
  3. Normalize and detrend the elasticities by dividing by the linear trends from step 2.
  4. Obtain country-specific parameters cIT,j and cCOM,j as averages of detrended series from step 3.
- Adopted specification (as presented):
  - αIT,jt = (0.0025    0.00022   )·IT_t + cIT,j
  - αCOM,jt = (0.006    0.00018   )·COM_t + cCOM,j
  - (These expressions imply that the average elasticity of output with respects to IT capital has grown from... [text continues in source beyond excerpt].)

*Source: Author’s calculations, as described in text (see discussion of steady-state analysis), based on data from IMF (2008), Global Insight (2006), and UN Statistics Division (2008).*

### 0.0025 in 1990 to 0.006 in 2006. Meanwhile, our estimates suggest that the average elasticity

### _wp1066 - 0.0025 in 1990 to 0.006 in 2006. Meanwhile, our estimates suggest that the average elasticity

### Elasticities of output with respect to ICT equipment
- The average elasticity of output with respect to IT equipment increased from 0.0025 in 1990 to 0.006 in 2006.
- The average elasticity of output with respect to communications equipment increased from 0.006 in 1990 to 0.009 in 2006.
- The R2 of the regressions estimating the linear trend in step 2 was 0.85 for the elasticity with respect to IT equipment and 0.47 for the elasticity with respect to communications equipment.

### Fluctuations in rate of decline of relative prices of ICT equipment
- Annual rates of price decline for ICT equipment fluctuate considerably:
  - For IT equipment: ranging from 8.8 percent to 26.4 percent.
  - For communications equipment: ranging from 5.6 percent to 10.7 percent.
- The immediate impact of a shock to relative prices of ICT equipment accounts for only 50 percent of the long-run effects.
- The subsequent adjustment is sluggish: half of the (remaining) adjustment takes about 13 years.
- Persistence of price-shock impacts and fluctuating rates of price decline imply that analyses based solely on steady-state properties may be misleading.

### Perturbation analysis framework and methodological notes
- Key equations referenced: Eq. (33) and Eq. (23), with (23) given as 22,    2,1 /  =  ttt pp.
- The approach interprets innovations in ICTs as shocks to the prices of ICT equipment and distinguishes immediate effects on capital stock from transitional effects as the economy reverts to its steady-state growth path.
- Small share of ICT equipment in the capital stock suggests limited changes to the average rate of depreciation and to the steady-state capital-output ratio; perturbation analysis assumes the steady-state capital-output ratio constant.
- Data availability constraints:
  - For IT equipment, price index for “computers and peripheral equipment” from the U.S. National Income and Product accounts is available from 1960.
  - For communications equipment, earlier price series are unavailable; a rate of price decline of 8 percent is applied for earlier years.
  - Elasticities of output with respect to ICT equipment for years before 1990 are taken as the 1990 estimates due to data limitations.
- Taking logs and running a panel regression with a time trend would yield similar results, but reliance on net imports data would require sample reduction or adjustments to handle occasional negative values.

### Estimated impacts of declining ICT equipment prices on growth (1990–2006)
- Aggregate findings:
  - Contributions of IT equipment and communications equipment to growth are similar, rising from about 0.09 percent at the beginning of the period to 0.13 towards the end.
  - On average, rates of investment in communications equipment are higher than for IT equipment (by a factor of about 2), but rates of price declines for IT equipment exceed those for communications equipment by a factor of about 2.2 over the 1990-2006 period, yielding similar growth effects.
- Period 1996–2000:
  - Contribution of capital deepening from declining prices of IT equipment peaks and exceeds the contribution from communications equipment.
  - This period aligns with earlier U.S.-focused literature on ICT impacts.
- Direct vs indirect effects:
  - Magnitude of direct and indirect effects is similar in steady state.
  - Year-to-year variations in growth impacts largely reflect direct effects of changing prices; indirect effects operate as a distributed lag with long memory (typical half-time about 13 years).
  - Indirect effects contribute about one-sixth to the acceleration of growth impacts of advances in ICTs between 1990 and 2006.
- Income-group differences:
  - Overall impacts are about one-third smaller in low-income countries compared to lower-middle-income countries.
  - Pattern of impacts over time is similar between low- and lower-middle-income countries due to similar weights of investment in IT equipment and model assumptions regarding parameter 2.
- Table 3 (selected country estimates, 1991-2006) — examples of total ICT equipment contributions (unweighted averages):
  - All countries covered: 0.19 (1991-1995), 0.25 (1996-2000), 0.26 (2001-2006).
  - Low-income countries: 0.15 (1991-1995), 0.20 (1996-2000), 0.21 (2001-2006).
  - Lower middle-income countries: 0.22 (1991-1995), 0.30 (1996-2000), 0.31 (2001-2006).
- Country examples from Table 3 (Total ICT Equipment):
  - Bangladesh: 0.07 (1991-1995), 0.10 (1996-2000), 0.10 (2001-2006).
  - China, P.R.: Mainland: n.a. entries for communications and total ICT Equipment; IT Equipment: 0.14 (1991-1995), 0.22 (1996-2000), 0.21 (2001-2006).
  - India: Total ICT Equipment: 0.18 (1991-1995), 0.23 (1996-2000), 0.25 (2001-2006).
  - Philippines: Total ICT Equipment: 0.30 (1991-1995), 0.39 (1996-2000), 0.43 (2001-2006).
  - Vietnam: Total ICT Equipment: 0.23 (1991-1995), 0.30 (1996-2000), 0.33 (2001-2006).
- Comparison of steady-state vs perturbation analysis:
  - Steady-state analysis sums direct and indirect impacts; perturbation analysis shows indirect effects occur slowly.
  - Elasticities used differ: steady-state calibrated elasticities reflect behavior in the calibration period and can be more affected by temporary accelerations in investment; perturbation analysis allows elasticities to change slowly over time (linear time trend), muting temporary swings.
  - Consequently, the gap between growth impacts of communications equipment between low- and lower-middle-income countries is smaller in the steady-state analysis than in the perturbation analysis.

### ICT production and contributions to growth
- Production-side contributions of productivity gains in ICT-producing sectors are conceptually calculated as the rate of productivity gains in the ICT-producing sector (indexed “2”), 22./AA, weighted by the sector’s share in the economy 2/YY.
- Data and measurement challenges:
  - Limited knowledge on inputs to ICT equipment production, particularly ICT-related components that embody technological advances.
  - Production data sometimes available only on a gross basis or lacking component-level input detail.
  - Difficulty distinguishing low-tech assembly operations from high-tech producers of electronic components.
- Mitigation strategy:
  - Include net exports of electronic microcircuits (SITC 2 category 7764) in estimates of ICT equipment production to capture microcircuit-related technological progress and to help distinguish assembly from component production; focusing on net exports avoids double-counting domestically produced microcircuits used in ICT equipment production.
- Price-based measure of productivity gains:
  - Use rate of decline of prices of ICT equipment (U.S. Producer Price Indices and U.S. National Income and Product Accounts); communications equipment deflators modified drawing on Doms (2005).
- Summary findings (Table 4 highlights):
  - In most countries covered, growth impacts of ICT equipment production are miniscule.
  - Growth impact of production of IT equipment exceeds 0.1 percentage points in only one low-income country (Vietnam) and 4 lower-middle-income countries.
  - In China, Philippines, and Thailand, impact of production of IT equipment on growth exceeds ½ percentage point at least in 2001-05.
  - Production of communications equipment has a less pronounced role: growth impact exceeds 0.1 percentage points in only two countries (India, Indonesia) in 2001-05, though it had larger impacts in the Philippines and Thailand in 1991-05.
  - Data constraints lead to exclusion of China from communications equipment production estimates; informal calculations suggest China’s production of communications equipment could also have a substantial growth impact.

*Source: Author’s calculations. See Appendix Table 2 for country-level estimates of all 88 countries covered by our analysis.*

### 0.2 percent in 2001-05.

### _wp1066 - 0.2 percent in 2001-05.

### Key quantitative findings
- China has been a net exporter of communications equipment since 2000, with net exports rising to about 1.5 percent of GDP by 2005.
- Including plausible values for domestic sales, domestic production of communications equipment in China could exceed 2 percent, consistent with a growth contribution of 0.2 percentage points or more.
- Philippines, 2001-05: the total contribution of production of ICT equipment to growth amounted to 2.2 percentage points, of which net exports of electronic microcircuits accounted for 1.5 percentage points.
- Between 1996 and 2006 the direct and indirect effects of falling prices of ICT equipment added about 0.2 percentage points to economic growth for low-income countries, and 0.3 percentage points for lower-middle-income countries.
- Relative to high-income countries, a preliminary analysis suggests:
  - Growth impacts of falling prices of IT equipment in low-income countries are about one-half of the level attained in high-income countries.
  - Growth impacts in lower-middle-income countries are about three-quarters of the level attained in high-income countries.

### Interpretation and caveats on measurement
- For some countries (notably Vietnam and China) net imports of electronic microcircuits account for a substantial proportion of the value of ICT equipment production, suggesting assembly of ICT equipment from imported components.
- Estimates rely on U.S. price indices for respective commodity categories; if the composition of production differs substantially from the commodity bundles underlying the U.S. indices, this introduces a margin of error.
- To relate estimated growth contributions from ICT production to official GDP growth requires knowledge of which deflators are applied in national accounts:
  - Only if national price indices fully reflect changes in the quality of ICT equipment is it possible to attribute a share of GDP growth to the estimated contributions from ICT-related production.
  - If national price indices do not capture changes in ICT product quality at all, it would be necessary to correct national GDP growth by adding the estimated growth contribution from productivity gains in the ICT-producing sector.
- A high contribution of ICT-related production to GDP does not necessarily translate into large national income gains because productivity gains can dissipate due to the terms-of-trade effect of falling ICT prices.

### Data construction and methodological notes
- A key obstacle is the lack of sufficiently detailed national accounts data for most low- and lower-middle-income countries.
- The study constructs estimates of investment in ICT equipment based on trade data and, where necessary, production data, compiling a database covering essentially all low- and lower-middle-income countries.
- Trade data are a noisy indicator for ICT-related spending, especially for small countries.
- An estimated mark-up is applied to “translate” trade data into spending figures in a national accounts framework; this adjustment adds noise because mark-ups likely vary across countries.
- The most important determinant of the link between falling ICT prices and growth is the elasticity of output with respect to ICT equipment:
  - Elasticities are estimated based on a steady-state version of the analytical framework, imposing equal rates of return to capital across asset types and using estimated investment data.
  - The study uses country fixed effects combined with a linear trend; alternative specifications could yield different elasticity estimates.
- Inclusion of net exports of electronic microcircuits refines findings about the growth impact of ICT production and highlights assembly-intensive production structures in several countries (Vietnam, China, Albania, India, Moldova, Sri Lanka, Ukraine).

### Sectoral patterns and structural observations
- Investment in communications equipment is about double the level of investment in IT equipment in both low- and lower-middle-income countries.
- Relative prices of IT equipment decline at a faster rate than communications equipment, so the overall contribution to growth from IT and communications equipment ends up broadly similar.
- Economic impacts of ICTs differ systematically across countries:
  - In high-income countries, spending on IT-related services plays a much higher role than in low-income countries.
  - ICTs have expanded tradable, labor-intensive services; investment in ICT equipment may not fully capture benefits for economies exporting such services.
  - Utilization of communication technologies requires both infrastructure investment and subscriptions to communication services; market structure and imperfect competition in telecommunications can affect how falling international ICT prices impact national economies.

### Limitations and outlook
- Trade-based estimates, mark-up adjustments, and reliance on U.S. price indices are important sources of uncertainty.
- Interpretation of changes in investment rates is challenging: they may be spurious, reflect business cycles, or indicate structural changes in elasticities.
- An explicit model deriving appropriate weights for elasticities is beyond the study’s scope; more model-based specifications could improve estimates.
- Broader economic impacts of ICTs extend beyond equipment absorption and include services, market structure effects, and tradability of services.

*Source: Author’s calculations and analysis as presented in the supplied document.*

### Appendix Table 1. Contribution of ICT-Related Capital Deepen

### Appendix Table 1. Contribution of ICT-Related Capital Deepening to Growth (Steady-State Analysis), 97 Countries, 2001-2006

### Overview
- Table reports investment shares (Percent of GDP), annual growth rates of IT Equipment and Communications Equipment, population growth (Annual Growth, in Percent), Real GDP Growth (Annual Growth, in Percent), elasticities of output w.r.t. IT Equipment and Communications Equipment, and the contribution to growth (Percentage Points) from IT Equipment, Communications Equipment, and ICT Equipment (=IT+C) for 97 countries over 2001-2006.
- Memorandum (all countries): Investment = 22.1 (Percent of GDP); IT Equipment growth = 0.74 (Annual Growth, in Percent); Communications Equipment growth = 1.39 (Annual Growth, in Percent); Population growth = 1.9 (Annual Growth, in Percent); Real GDP Growth = 5.1 (Annual Growth, in Percent); Elasticity of output w.r.t. IT Equipment = 0.006; Elasticity of output w.r.t. Communications Equipment = 0.007; Contribution to Growth — IT Equipment = 0.162 (Percentage Points); Communications Equipment = 0.139 (Percentage Points); ICT Equipment (=IT+C) = 0.301 (Percentage Points).

### Selected country contributions (examples extracted from the table)
- Afghanistan, I.R. of:
  - Investment = 39.4 (Percent of GDP); IT Equipment growth = 0.62; Communications Equipment growth = 2.67; Population growth = 4.1; Real GDP Growth = 11.7; Elasticities = 0.004 (IT), 0.010 (Communications); Contributions = 0.12 (IT Equipment), 0.20 (Communications Equipment), 0.31 (ICT Equipment).
- Paraguay:
  - Investment = 18.6 (Percent of GDP); IT Equipment growth = 5.62; Communications Equipment growth = 1.98; Population growth = 2.0; Real GDP Growth = 2.9; Elasticities = 0.042 (IT), 0.011 (Communications); Contributions = 1.21 (IT Equipment), 0.21 (Communications Equipment), 1.42 (ICT Equipment).
- Djibouti:
  - Investment = 17.8 (Percent of GDP); IT Equipment growth = 1.68; Communications Equipment growth = 3.59; Population growth = 1.9; Real GDP Growth = 3.1; Elasticities = 0.013 (IT), 0.019 (Communications); Contributions = 0.38 (IT Equipment), 0.36 (Communications Equipment), 0.74 (ICT Equipment).
- Maldives:
  - Investment = 32.3 (Percent of GDP); IT Equipment growth = 2.28; Communications Equipment growth = 5.70; Population growth = 1.6; Real GDP Growth = 6.9; Elasticities = 0.017 (IT), 0.027 (Communications); Contributions = 0.48 (IT Equipment), 0.51 (Communications Equipment), 0.98 (ICT Equipment).
- China, P.R.: Mainland:
  - Investment = 39.3 (Percent of GDP); IT Equipment growth = 0.46; Communications Equipment data unavailable; Population growth = 0.7; Real GDP Growth = 9.8; Elasticity of output w.r.t. IT Equipment = 0.003; Contribution to Growth — IT Equipment = 0.10 (Percentage Points); Communications Equipment contribution data unavailable.
- India:
  - Investment = 27.1 (Percent of GDP); IT Equipment growth = 0.88; Communications Equipment growth = 1.76; Population growth = 1.6; Real GDP Growth = 7.0; Elasticities = 0.006 (IT), 0.009 (Communications); Contributions = 0.19 (IT Equipment), 0.16 (Communications Equipment), 0.35 (ICT Equipment).
- Paraguay and other high IT-contribution examples illustrate that large IT Equipment growth rates and high elasticities can produce large contributions to GDP growth (e.g., Paraguay IT Equipment contribution = 1.21 Percentage Points).

### Regional / income-group aggregates (Memorandum items)
- LIC (Low-Income Countries):
  - Investment = 20.2 (Percent of GDP); IT Equipment growth = 0.6; Communications Equipment growth = 1.3; Population growth = 2.5; Real GDP Growth = 4.8; Elasticities = 0.005 (IT), 0.007 (Communications); Contributions = 0.131 (IT Equipment), 0.128 (Communications Equipment), 0.259 (ICT Equipment).
- LMC (Lower-Middle-Income Countries):
  - Investment = 23.8 (Percent of GDP); IT Equipment growth = 1.0; Communications Equipment growth = 1.5; Population growth = 1.4; Real GDP Growth = 5.3; Elasticities = 0.007 (IT), 0.008 (Communications); Contributions = 0.213 (IT Equipment), 0.151 (Communications Equipment), 0.364 (ICT Equipment).

### Key empirical patterns visible in the table
- Across the 97-country sample (Memorandum), ICT Equipment contributed 0.301 (Percentage Points) to annual GDP growth on average during 2001-2006, with IT Equipment contributing 0.162 and Communications Equipment contributing 0.139.
- Elasticities of output w.r.t. IT Equipment and Communications Equipment are small in magnitude (for the full sample: 0.006 and 0.007, respectively), yet nonzero contributions to growth arise because of positive investment shares and equipment growth rates.
- Country-level contributions vary markedly: examples range from small ICT contributions (e.g., Yemen, Republic of: ICT Equipment = 0.09 Percentage Points) to large contributions (e.g., Paraguay: ICT Equipment = 1.42 Percentage Points; Djibouti: ICT Equipment = 0.74 Percentage Points).

### Notes and data caveats reported in the table
- Footnote 1: Investment rates were unavailable for some countries. Elasticities and contributions of ICT equipment to growth were estimated based on the sample average for investment rates (22.0 percent).
- Footnote 2: For Afghanistan, data relate to 2002-2006.
- Footnote 3: For China, data on investment in communications equipment were unavailable.
- Footnote 4: Totals exclude China (as estimates on investment in communications equipment were unavailable) and Paraguay (owing to severe inconsistencies in published trade data).

*Source: Authors calculations, as described in text, based on data from IMF (2008), Global Insight (2006), and UN Statistics Division (2008).*

### Appendix Table 2. Contribution of ICT-Related Capital Deepening to Growth, 88 Countries, 1990-2006

### Appendix Table 2. Contribution of ICT-Related Capital Deepening to Growth, 88 Countries, 1990-2006

### Key findings from the table (country-level contributions)
- The table reports contributions to growth (in percentage points) from IT Equipment, Communications Equipment, and Total ICT Equipment for three periods: 1991-1995, 1996-2000, and 2001-2006.
- Selected country examples (Total ICT Equipment contributions):
  - Maldives: 0.51 (1991-1995), 0.66 (1996-2000), 0.72 (2001-2006)
  - Guyana: 0.40 (1991-1995), 0.54 (1996-2000), 0.56 (2001-2006)
  - Macedonia, FYR: 0.34 (1991-1995), 0.46 (1996-2000), 0.47 (2001-2006)
  - Philippines: 0.30 (1991-1995), 0.39 (1996-2000), 0.43 (2001-2006)
  - Djibouti: 0.35 (1991-1995), 0.46 (1996-2000), 0.48 (2001-2006)
  - China, P.R.: Mainland: IT Equipment: 0.14 (1991-1995), 0.22 (1996-2000), 0.21 (2001-2006); Communications Equipment and Total ICT Equipment: n.a. (not available)
- Many low- and lower-middle-income countries show rising contributions from both IT Equipment and Communications Equipment across the three periods; examples include Cape Verde (Total: 0.29, 0.40, 0.41) and Tonga (Total: 0.32, 0.43, 0.45).

### Country-group (unweighted average) results
- All countries covered (unweighted averages):
  - IT Equipment: 0.09 (1991-1995), 0.14 (1996-2000), 0.13 (2001-2006)
  - Communications Equipment: 0.09 (1991-1995), 0.11 (1996-2000), 0.13 (2001-2006)
  - Total ICT Equipment: 0.19 (1991-1995), 0.25 (1996-2000), 0.26 (2001-2006)
- Low-income countries (LIC):
  - IT Equipment: 0.08 (1991-1995), 0.12 (1996-2000), 0.11 (2001-2006)
  - Communications Equipment: 0.07 (1991-1995), 0.08 (1996-2000), 0.10 (2001-2006)
  - Total ICT Equipment: 0.15 (1991-1995), 0.20 (1996-2000), 0.21 (2001-2006)
- Lower-middle-income countries (LMC):
  - IT Equipment: 0.11 (1991-1995), 0.17 (1996-2000), 0.16 (2001-2006)
  - Communications Equipment: 0.11 (1991-1995), 0.13 (1996-2000), 0.15 (2001-2006)
  - Total ICT Equipment: 0.22 (1991-1995), 0.30 (1996-2000), 0.31 (2001-2006)

### Data appendix: production and data sources (methodology notes)
- Primary production data source: UNIDO Industrial Statistics Database (UNIDO, 2007), ISIC categories used:
  - ISIC 3 category 3000 (“manufacture of office, accounting, and computing machinery”) and ISIC 2 category 3825 for IT equipment.
  - ISIC 3 category 30 (“manufacturing of radio, television, and communication equipment and apparatus”) for communications equipment.
- Coverage limitations and adjustments:
  - Limited UNIDO coverage for many low-income and lower-middle-income countries prompted complementary analysis of trade data (United Nations Statistics Division, 2008) to identify additional producers.
  - Three producers identified outside UNIDO coverage: China, Thailand, and Tunisia.
  - For Indonesia, UNIDO (2007) shows trivial IT production despite significant net exports since 1994; estimates rely on trade data validated by partner countries.
  - For Macedonia, UNIDO (2007) production data appeared inconsistent with negligible exports (less than 0.02 percent of GDP); estimates for Macedonia follow trade data, assuming production equal to zero.
- Data construction details:
  - Appendix Tables 3 and 4 summarize production data used for 1990-2005.
  - Gaps in annual UNIDO observations were filled by intrapolation (data in italics in the source).
  - For China and Indonesia, production of ICT equipment is based primarily on trade data from United Nations Statistics Division (2008) and domestic spending data from WITSA (2001) and Global Insight (2006).

### Appendix Table 3 — selected production figures and memorandum items (1990-2005)
- Representative production datapoints (IT Equipment, percent of GDP) reported in Appendix Table 3:
  - India: 1990: 0.20; 1991: 0.28; 1992: 0.30; 1994: 0.31; 1995: 0.29; 2000: 0.22; 2005: 0.24
  - China, P.R.: Mainland: 1990: 0.26; 1991: 0.26; 1992: 0.32; 1995: 0.52; 1999: 0.98; 2000: 1.31; 2003: 3.39; 2004: 4.33; 2005: 4.70
  - Indonesia: 1990: 0.03; 1991: 0.08; 1992: 0.15; 1995: 0.32; 1998: 0.70; 2000: 1.95; 2001: 1.42; 2005: 1.01
  - Philippines: 1990: 0.02; 1993: 0.40; 1994: 0.47; 1998: 3.77; 1999: 4.21; 2000: 4.37; 2005: 3.66
  - Vietnam: increasing notable production share in later years reported (examples include 2002: 1.14; 2003: 1.37; 2004: 1.60; 2005: 0.53 in the source table layout)
- Memorandum item — assumed share of IT production in ISIC Rev. 3 category 3000 (unweighted average used for scaling):
  - 78.6, 77.9, 75.2, 82.5, 77.0, 85.4, 85.4, 86.2, 87.2, 86.4, 88.5, 86.4, 89.5, 90.2, 89.9, 89.8 (across years shown)

*Source: Author’s calculations, based on UNIDO (2007), United Nations Statistics Division (2008), WITSA (2001), Global Insight (2006), and IMF (2008), as presented in the source content.*

### Appendix Table 4. Production of Communicati

### Appendix Table 4. Production of Communications Equipment in Low- and Lower-Middle Income Countries, 1990-2005 (Percent of GDP)

### Production observations (selected country series shown in table)
- India: 0.80 (1990), 0.84 (1991), 0.99 (1992), 0.83 (1993), 1.05 (1994), 1.02 (1995), 0.90 (1996), 1.11 (1997), 1.04 (1998), 1.00 (1999), 0.98 (2000), 0.82 (2001), 1.18 (2002), 1.06 (2003), 1.06 (2004), 1.06 (2005)
- Kyrgyz Republic: 0.00 (1990), 0.00 (1991), 0.00 (1992), 0.00 (1993), 0.00 (1994), 0.00 (1995), 0.00 (1996), 0.02 (1997), 0.02 (1998), 0.15 (1999), 0.15 (2000), 0.17 (2001), 0.19 (2002), 0.10 (2003), 0.18 (2004), 0.18 (2005)
- Vietnam: 0.00 (1990), 0.00 (1991), 0.00 (1992), 0.00 (1993), 0.00 (1994), 0.00 (1995), 0.00 (1996), 0.00 (1997), 1.54 (1998), 1.58 (1999), 1.63 (2000), 0.00 (2001), 0.00 (2002), 0.00 (2003), 0.00 (2004), 0.00 (2005)
- Zimbabwe: 0.21 (1990), 0.22 (1991), 0.20 (1992), 0.16 (1993), 0.19 (1994), 0.20 (1995), 0.14 (1996), 0.00 (1997), 0.00 (1998), 0.00 (1999), 0.00 (2000), 0.00 (2001), 0.00 (2002), 0.00 (2003), 0.00 (2004), 0.00 (2005)
- Albania: 0.00 (1990), 0.00 (1991), 0.00 (1992), 0.11 (1993), 0.00 (1994), 0.06 (1995), 0.00 (1996), 0.00 (1997), 0.00 (1998), 0.00 (1999), 0.00 (2000), 0.00 (2001), 0.00 (2002), 0.00 (2003), 0.00 (2004), 0.00 (2005)
- Egypt: 0.16 (1990), 0.24 (1991), 0.34 (1992), 0.31 (1993), 0.28 (1994), 0.41 (1995), 0.35 (1996), 0.29 (1997), 0.36 (1998), 0.33 (1999), 0.30 (2000), 0.28 (2001), 0.25 (2002), 0.25 (2003), 0.25 (2004), 0.25 (2005)
- Indonesia: 0.49 (1990), 0.43 (1991), 0.86 (1992), 0.81 (1993), 1.06 (1994), 1.41 (1995), 1.65 (1996), 2.16 (1997), 2.35 (1998), 2.16 (1999), 3.11 (2000), 1.59 (2001), 1.72 (2002), 1.96 (2003), 1.96 (2004), 1.96 (2005)
- Iran, I.R. of: 0.34 (1990), 0.34 (1991), 0.34 (1992), 0.35 (1993), 0.40 (1994), 0.23 (1995), 0.39 (1996), 0.65 (1997), 0.89 (1998), 1.26 (1999), 1.65 (2000), 1.47 (2001), 0.47 (2002), 0.39 (2003), 0.39 (2004), 0.39 (2005)
- Moldova: 0.00 (1990), 0.00 (1991), 0.00 (1992), 0.00 (1993), 0.00 (1994), 0.00 (1995), 0.65 (1996), 0.34 (1997), 0.00 (1998), 0.00 (1999), 0.00 (2000), 0.00 (2001), 0.00 (2002), 0.00 (2003), 0.15 (2004), 0.15 (2005)
- Morocco: 0.00 (1990), 0.00 (1991), 0.00 (1992), 0.00 (1993), 0.00 (1994), 0.00 (1995), 0.16 (1996), 0.21 (1997), 0.20 (1998), 0.02 (1999), 0.40 (2000), 0.23 (2001), 0.23 (2002), 0.25 (2003), 0.28 (2004), 0.28 (2005)
- Peru: 0.31 (year not specified in table fragment), 0.28 (year not specified), 0.42 (year not specified), 0.24 (year not specified), 0.05, 0.03, 0.03, 0.03, 0.03, 0.03, 0.03, 0.03, 0.03 (sequence shown in table fragment)
- Philippines: 3.07 (1990), 3.92 (1991), 4.08 (1992), 3.78 (1993), 4.15 (1994), 4.87 (1995), 0.00 (1996), 0.00 (1997), 0.00 (1998), 0.00 (1999), 0.00 (2000), 0.00 (2001), 0.00 (2002), 0.00 (2003), 0.00 (2004), 0.00 (2005)
- Sri Lanka: 0.07 (1990), 0.13 (1991), 0.07 (1992), 0.03 (1993), 0.10 (1994), 0.10 (1995), 0.10 (1996), 0.08 (1997), 0.10 (1998), 0.13 (1999), 0.11 (2000), 0.09 (2001), 0.09 (2002), 0.09 (2003), 0.09 (2004), 0.09 (2005)
- Thailand: 3.51 (year indicated in table fragment), 6.62 (year indicated), 5.87 (year indicated), 5.13, 2.62, 3.51, 4.41, 2.20, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00 (sequence shown in table fragment)
- Ukraine: 0.00 (1990), 0.00 (1991), 0.00 (1992), 0.00 (1993), 0.00 (1994), 0.00 (1995), 0.00 (1996), 0.00 (1997), 0.00 (1998), 0.00 (1999), 0.41 (2000), 0.54 (2001), 0.49 (2002), 0.51 (2003), 0.66 (2004), 0.66 (2005)
- Note: China, P.R.: Mainland entries in the table are marked "n.a." for the years shown in the fragment.

### Data on trade in ICT equipment (methods and coverage)
- Primary source for constructing absorption dataset: United Nations Statistics Division (COMTRADE database) (2008).
- Principal SITC 2 categories used:
  - IT equipment: 752 (automatic data processing equipment) and 7599 (parts and accessories pertaining equipment in category 752), corresponding to HS 2002 categories 8471 and 847330.
  - Electronic microcircuits (for some analyses): SITC 2 category 7764.
  - Communications equipment: SITC 2 category 764 (telecommunication equipment, parts and accessories).
- Where national trade data are unavailable, proxies are constructed by adding corresponding trade flows from reporting countries and scaling to account for the share of non-reporting countries in recorded trade (see Fig. A1 reference).
- Dataset constructed using:
  - Data as reported by countries of interest where available; and
  - For non-reporting countries, proxies from reporting countries scaled to account for share of non-reporting countries in recorded trade.

### Spending on ICT equipment (conversion from trade to expenditure)
- Trade data used are on a fob basis and exclude insurance, freight, customs clearance, import tariffs, indirect taxes (value added tax, sales tax, excise taxes), retail costs, and profit margins.
- Mark-up to convert net imports into spending:
  - On average, spending on IT hardware from Global Insight (2006) exceeds net imports by 72 percent. The study adopts a mark-up of 72 percent for both IT equipment and communications equipment.
- Rules applied to estimate spending:
  - For IT equipment: use Global Insight (2006) and WITSA (2001) where available; otherwise apply a mark-up of 72 percent to net imports in SITC categories 752 and 7599.
  - For communications equipment: where domestic spending figures are unavailable, apply a mark-up of 72 percent to imports in SITC 2 category 764.
  - For producers of ICT equipment:
    - Use Global Insight (2006) and WITSA (2001) where available for IT equipment spending.
    - For communications equipment, or IT equipment where spending data are unavailable, apply 72 percent mark-up to net imports.
    - Estimate total ICT-related spending as the sum of relevant net imports (plus mark-up) and estimated production levels at face value.
- Countries for which Global Insight (2006) spending data were available or classified as producers (list): China, Egypt, India, Indonesia, Iran, Morocco, Philippines, Sri Lanka, Thailand, and Ukraine.
- Countries identified as non-producers used for mark-up comparison (14 countries): Bangladesh, Bolivia, Cameroon, Colombia, Ecuador, Honduras, Jamaica, Jordan, Kenya, Pakistan, Peru, Senegal, Tunisia, Zimbabwe.
- Caveat noted: a zero mark-up is applied to domestic production values in the aggregation to account for potential intermediate inclusion in net import data and to avoid double-counting.

### Prices of ICT equipment (relative price indices, 1990=100)
- General approach: decline in relative prices of ICT equipment based on U.S. national accounts data and PPI series, with communication equipment series following Doms (2005) for improved measurement.
- Appendix Table 5 — Relative Prices of ICT Equipment, 1990-2006 (1990=100):
  - Columns: NIPA: Investment: Computers and peripheral equipment; PPI: Communication and related equipment; Communication and related equipment (based on Doms, 2005); PPI: Integrated microcircuits
  - 1990: 100.0 ; 100.0 ; 100.0 ; 100.0
  - 1991: 86.7 ; 97.7 ; 91.8 ; 85.3
  - 1992: 71.9 ; 95.4 ; 84.3 ; 73.4
  - 1993: 59.7 ; 94.3 ; 78.3 ; 70.1
  - 1994: 51.4 ; 93.8 ; 73.2 ; 69.7
  - 1995: 42.1 ; 92.9 ; 68.4 ; 61.1
  - 1996: 31.5 ; 91.6 ; 64.6 ; 46.4
  - 1997: 23.9 ; 90.9 ; 59.8 ; 34.4
  - 1998: 17.6 ; 90.2 ; 54.7 ; 25.0
  - 1999: 13.4 ; 87.6 ; 49.5 ; 22.7
  - 2000: 11.4 ; 83.9 ; 44.2 ; 20.3
  - 2001: 9.2 ; 81.4 ; 40.2 ; 16.1
  - 2002: 7.8 ; 78.8 ; 36.6 ; 14.3
  - 2003: 6.8 ; 75.8 ; 33.0 ; 13.6
  - 2004: 6.2 ; 72.3 ; 29.5 ; 12.5
  - 2005: 5.3 ; 69.7 ; 26.6 ; 11.5
  - 2006: 4.5 ; 67.7 ; 24.3 ; 10.4
- Note on Doms (2005) series: estimates from Doms (2005) were available for 1994-2000 only. For 1990-93 and 2001-06 (numbers shown in italics in the original table), the series were extrapolated assuming that the price index for communication equipment declines at a rate that is 6 percent faster than in the PPI statistics.

### Other data and country coverage
- GDP (millions of U.S. dollars) and GDP per capita (U.S. dollars) obtained from IMF’s World Economic Outlook database (IMF, 2008).
- Country groupings (low-, lower-middle-, upper-middle-, and high-income) based on World Bank’s country classification for 2007.
- Trade dataset coverage: includes all 49 countries classified as low-income and all 54 countries classified as low-middle-income; analysis largely confined to a group of 95 countries for which a minimum set of data are available from IMF (2008).
- Countries excluded from the main analysis due to data limitations: Democratic People's Republic of Korea, Liberia, Somalia, Federated States of Micronesia, Iraq, Marshall Islands, West Bank and Gaza, and Timor-Leste.

*Source: Author’s calculations, based on data from UNIDO (2007) and IMF (2008); additional sources cited in text (United Nations Statistics Division, 2008; Global Insight (2006); WITSA (2001); Doms (2005); U.S. Department of Commerce, Bureau of Economic Analysis (2008); U.S. Department of Labor, Bureau of Labor Statistics (2008)).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1066.pdf_