Mrkkkk

These tools are widely used by local economic development (LED) practitioners:
General tools to help organise and compare data: Time series analysis Growth indexes Composite indexes Benchmarking GIS mapping PEST / trends analysis Tools to help cities understand the structure of their local economy: Sector share analysis Value-added analysis Economic base analysis Location quotient Specialisation index Shift share analysis Input-output analysis Social accounting matrix Cluster mapping Value chain analysis Tools to look at local endowments: Asset mapping Tools to assess human capital: Skills audit Tools to analyse institutions: Stakeholder analysis / institutional mapping

Analysing the Data
Guide to Data Analysis Tools

8

This chapter discusses the tools that cities and city-regions can use to analyse data collected on their economies. All of these have been tried and tested in actual city development strategies (CDSs). A few other tools not yet widely applied in city planning environments have also been included.

Time Series Analysis
What Issues Are Addressed by Time Series Analysis?
The following questions can be addressed by a time series analysis: How is a local economy performing over time? ● Population and other demographics (including education and labour force ● Income levels and distribution ● Employment and unemployment levels (total economy and by sector) ● Economic output and exports (total economy and by sector) Which growth patterns reﬂect shocks and cycles and which are long-term trends?

73

How Is Time Series Analysis Used?

See table 4.1, page 36

Time series analysis is one of the most widely used tools in local economy assessments. It maps the development of key socioeconomic indicators over a period of time and displays them in tables and graphs (line graphs or bar charts). Changes over time are expressed through annual growth percentages and compound annual growth rates (CAGR) and growth indexes. One of the great advantages of this analysis versus more static snapshot type tools is that it enables cities to determine whether an indicator for a speciﬁc year is a shock or the result of a long-term trend. Table 8.1 is a basic example of the use of time series analysis for looking at local employment levels (by sector) in a basic three-sector economy.

ANALYSING THE DATA

T A B L E 8 . 1 Time Series Analysis for Local Employment Levels in a Three-Sector Economy
Sector
Agriculture Manufacturing Services TOTAL

2002 %
2,009 11,350 14,500 27,859 — — — —

2003 %
2,000 11,600 15,330 28,930 – 0.4 +2.2 +5.7 +3.8

2004
1,980 11,670 16,150 29,800

%
– 1.0 +0.6 +5.3 +3.0

2005 %
1,974 11,145 16,553 29,652 – 0.3 – 4.5 +2.4 – 0.5

CAGR (2002–2005)
– 0.6 – 0.6 +4.5 +2.1

LED RESOURCE GUIDE

CASE STUDY
MUNICH (Germany)
Time series analysis is one of the most frequently used tools in Munich’s annual local economy assessment. It is used to analyse and present trends over time for key LED indicators, including employment levels, sector growth, and unemployment levels. Although time series data are typically analysed on a ﬁve-year cycle, analyses of medium-term trends (over periods of 10 to 20 years) are also conducted. Munich places particular emphasis on benchmarking its performance (by comparing its performance with other German and international cities), so it analyses and presents comparable historical data for benchmark cities as well. Munich presents the ﬁndings from its time series analysis in several ways, including bar charts, graphs, and tables. And the visual presentation of ﬁndings is accompanied by a clear narrative explanation of trends and the main conclusions to be drawn from the analysis.

74

In addition to tables, time series analysis is normally presented using graphics, most often trend lines, histograms, and bar charts. Time series analysis is often used in combination with sector share analysis, value-added analysis, and benchmarking,

and is fundamental to developing growth indexes. In particular, the trends of the key indicators in question are often tracked against national trends, since this may reveal whether, for example, a recession is a local phenomenon or a reﬂection of a wider trend experienced at the national level.

Tips
Compare with regional/national/ trends. Time series analysis can be especially helpful when comparing local trends with regional and/or national ones. However, because of the differing size of the economies surveyed, the data are not always directly comparable, so it is useful to construct a growth index (see page 75) to aid in comparisons. Complement the data with insights from experts and literature. Time series analysis displays trends for the economic indicators in question but does not explain or analyse them, and it does not provide predictive power. By consulting relevant experts and literature in addition to conducting a time series analysis, cities may be able to identify the factor(s) that explain, for example, a sudden change in a particular sector (decline in demand, increased competition, or rapid structural changes) and be in a better position to assess the future competitiveness of the local economy. Making the best use of visual tools. Spreadsheet software like Microsoft Excel provides valuable tools that can be used to plot graphs for time series analysis. Visual tools can be powerful and communicate information effectively. However, keep the target audience in mind and develop graphs and explanatory text appropriately.

Further Information
For guidelines to (and an example of) the use of time series analysis in the United States, see: http://www.economictoolbox.geog. psu.edu/index.php. For guidelines on how to display time series data in a line graph, see: http://erc.msh.org/quality/foutools/foulngrf.cfm. For a discussion of time series analysis as part of a forecasting exercise, see: http://www.statsoft.com/textbook/stathome.html. For a discussion on the use of time series analysis in forecasting, see chapter 4 in Regional Economic Modeling: A Systematic Approach to Economic Forecasting and Policy Analysis, by George I. Treyz (1993, Kluwer Academic Publishers).

What Key Inputs Are Required for Time Series Analysis?
This analysis requires data for the selected socioeconomic indicator (including employment, gross domestic product [GDP], and population size) over several years. Time series analysis is easy to conduct and does not require any knowledge of econometrics beyond basic statistical analysis. This analysis does not require any special resources other than data and human resources, but spreadsheet software with the capability to plot graphs is useful. When comparable national and local level data are available, and other data collection is not required, the analysis has a low resource intensity.

The growth index converts absolute data (including employment, output, and productivity) in a reference year for any number of economies into a common value (normally 100). This enables simple comparisons of relative performance, particularly for absolute values that differ substantially. To calculate a growth index, set the “year n” at 100; for each subsequent year the formula is then: Year n + 1 index = (year n + 1 / year n) x 100; Year n + 2 index = (year n + 2 / year n + 1) x (year n + 1), etc. These values can then be presented in a tabular format (see table 8.2) and also using line charts and bar charts. An advantage of using indexes to compare local indicators with regional or national indicators is that an index may reveal, for example, whether an economic phenomenon (like recession) represents a wider trend experienced on a national level or if it is a more local event.

75

Growth Indexes
What Issues Are Addressed by Growth Indexes?
The following questions can be addressed by growth indexes: How do various aspects of the local economy’s performance compare with other economies over time? Which local growth patterns are driven by shocks or cycles? And which patterns are long-term trends?

What Key Inputs Are Required for Growth Indexes?
This type of analysis requires at least two years of time series data, for the local economy and for a reference economy

How Are Growth Indexes Used?
The growth index is one of the most commonly used tools in analysing the local economy. It is a simple and cost-effective way to measure and compare local economic performance with that of another economy (or economies), because it allows for direct comparison of a particular socioeconomic indicator between two or more economies over time. The growth index is often used to apply benchmarking comparisons to a basic time series analysis.

T A B L E 8 . 2 An Example of the Presentation of Data Using the Growth Index
FDI Economy 1 FDI Economy 2 Index–FDI 1 Index–FDI 2
2002 2003 2004 2005 50 60 68 81 350 390 440 470 100 120 136 162 100 111 126 134

FDI: foreign direct investment

ANALYSING THE DATA

CASE STUDY
RAFAELA (Argentina)
The city of Rafaela calculated a growth index to make comparisons across ﬁrms of different sizes (based on the number of employees, fewer than 5, 6–10, 11–20, 21–50, 51–100, and more than 100 ) over a period of four years (1997–2001). The number of ﬁrms per category was converted into a common value of 100 for the reference year (in this case 1997). The employment growth trend was then tracked for the next three years, based on the reference year, and presented in a histogram. By using the growth index, Rafaela was able to compare employment growth across categories of ﬁrms—despite the fact that the number of ﬁrms in each category varied substantially during this time period.

LED RESOURCE GUIDE

Tips
Complement the data with insights from experts and literature. Growth indexes display trends for the indicators in question but do not explain or analyse them. And these indexes do not have predictive power. By consulting relevant experts and literature, in addition to a time series analysis, it may be possible to identify the factor(s) that drive a sudden change in a particular sector (increased competition, rapid structural changes, or decline in demand, for example)—which would help the city to assess the likely future competitiveness of the local economy. Making the best use of visual tools. Spreadsheet software like Microsoft Excel offers valuable tools for plotting graphs of growth indexes. These visual tools can be powerful and can often communicate information more clearly than detailed data tables. But graphs can also be misused and (in some cases) make information less clear (as is sometimes true for three-dimensional bar charts, for example). In general, simple graphs with clear explanatory text work best.

Further Information
For guidelines on constructing a growth index, and an example of its use in the United States, see: http://www.eco nomictoolbox.geog. psu.edu/index.php. For guidelines and an illustration on how to construct and use a growth index, see: http://www.oecd.org/home/ 0,2987,en_2649_201185_1_1_1_1_1,00.html. For an example of how growth indexes are used in a benchmarking analysis by the City of Glasgow (Scotland), see Glasgow Economic Analysis and Benchmarking Study 2005 (beginning at page 14): http://www.glasgoweconomicfacts. com/Glasgow%20Report%20BAK%20-%20 Executive%20Summary%20Des%204.pdf.

76

(often the national economy). Data availability depends on the type of data to be indexed. Historical data on employment and population tend to be more readily available than city-level GDP data, for example. This analysis does not require econometric knowledge. It can be conducted using basic spreadsheet software, so this tool is low in resource intensity.

titioners to summarise complex phenomena with a single indicator.

How Are Composite Indexes Used?
A composite indicator aggregates a set of indicators to construct a single measurement of a socioeconomic phenomenon. Composite indicators are used extensively in social science research by governments, international organisations, and universities to measure complex economic events. In particular, they are often used to assess investment climate as in the World Bank Ease of Doing Business Index; see investment climate survey, page 66) and to analyse poverty

Composite Indexes
What Issues Are Addressed by Composite Indexes?
This tool can be used to measure any socioeconomic issue relevant to a local economy assessment and it allows prac-

Skyline view of Medellin, Colombia

CASE STUDY
MEDELLÍN (Colombia)
In 2001 the city of Medellín created a quality of life (QoL) index to assess social exclusion and poverty in the city—and, since 2004, has monitored this index on an annual basis. The data for the QoL index, collected through a survey of 20,000 households, are used to measure the geographical distribution of poverty and identify areas where policies to combat social exclusion should be targeted. Out of more than 50 indicators used to measure the quality of housing, access to public services and social security, asset endowment of the household, and level of schooling of adults and children in the household, 17 indicators were selected as the base for the QoL index. The composite indicator e Im
W age

oth Ti m

h ss/T y Ro

and deprivation (as in the UN-HABITAT City Development Index and UNDP’s Human Development Index). The advantage of using this approach rather than a set of individual indicators is that a composite indicator provides a better picture of complex issues and simpliﬁes the rankings and comparisons. Composite variables are also seen as a particularly effective way to communicate trends of interest to policy makers and local stakeholders. Calculating composite indicators often involves using a group of statistical methods called data reduction techniques. Two of the more common techniques are principal component analysis, which identiﬁes groups of indicators whose scores (or behaviour) are driven by the same underlying factor, and unobserved component analysis, which removes outlying indicators. A simpler way to construct composite indicators is to weight individual indicators according to importance and add the results. Table 8.3 shows an example of how the compos-

ite index (total adjusted score) would be calculated using this method. (Based on a simple composite of three factors, the weighting of the factors should add up to 1.0.)

77

Tips
Consider costs and beneﬁts. Oversimpliﬁcation can be a problem when using composite indicators, because socioeconomic situations are often too complicated to be adequately captured within a single index. In some cases, presenting several key indicators (possibly in addition to a composite) may give a more balanced picture. Select the right indicators. A composite index is only as good as the indicators used to construct it. Therefore, it is important to think and assess carefully before aggregating indicators into a composite. For example, do the indicators measure elements of the same phenomenon? How might the indicators be interrelated? To test a composite index, replace a key indicator with an alternate indicator and make sure the results are not radically different. Be careful with applying weights. From a technical standpoint, applying weights is a simple process. But the process can be politically difﬁcult since results may be highly sensitive to the weights applied. So it is important to have a clear methodology in place for determining how to weight various factors, and the weighting process should be developed and tested with key experts and stakeholders.

T A B L E 8 . 3 Calculation of Composite Index Using a Weighting Method
Factor 1
Raw score Weighting Adjusted score 5.0 0.3 1.5

Factor 2
7.0 0.1 0.7

Factor 3
9.0 0.6 5.4

Total
21.0 1.0 7.6

ANALYSING THE DATA

was calculated using principal component analysis.The resulting index value ranges from 0 to 100 (with the lower scores indicating a lower quality of life). For Medellín, the main advantage of using a QoL index is the ability to account for the many factors that deﬁne poverty in one indicator.

ork

s

Further Information
For guidelines and an example of how to create a composite variable, see: http://www.doingbusiness.org/documents/how_to_ aggregate.pdf. For guidelines on how to use the principal component analysis, see (under factor analysis): http://www.statsoft.com/textbook/ stathome.html. For guidelines on calculating the United Nations Development Programme’s (UNDP’s) Human Development Index, see: http://hdr. undp.org/docs/statistics/indices/technote_1.pdf. For a critical review of the use of composite indicators in city rankings, see Grading Places: What Do the Business Climate Rankings Really Tell Us?, by Peter Fisher: http://www.epinet.org/content.cfm?id=2052.
LED RESOURCE GUIDE

For an in-depth discussion and guidelines on calculating indexes, see State-of-the-Art Report on Composite Indicators for the Knowledgebased Economy, prepared by the Sixth Framework Programme of the European Commission: http://kei.publicstatistics.net/ KEI%20D5_1.pdf. For comprehensive guidance on composite variables, see the OECD’s Handbook on Constructing Composite Indicators: Methodology and User Guide: http://www.olis.oecd.org/olis/2005doc.nsf/LinkTo/std-doc(2005)3. For more information and links to a wide range of socioeconomic indexes (mostly calculated on a national basis), see: http://human development.bu.edu/use_exsisting_index/start_content.cfm.

78

The composite indicator is often tracked over time using time series analysis.

Benchmarking
What Issues Are Addressed by Benchmarking?

What Key Inputs Are Required for Composite Index Studies?
Because composite indexes are normally only constructed to simplify complex data, access to fairly extensive data is a requirement for using this tool. Such data may come from public statistical sources or may require the use of surveys (an investment climate survey, for example), depending on the information to be analysed. For basic composite indicators that use a weighting technique, the analysis is relatively simple; it requires no econometric analysis and only a basic understanding of statistics. However, if the city plans to make extensive use of composite indexes, it will be necessary to have an understanding of more complex data reduction techniques. Statistical software packages like SPSS or STATA would be valuable to support the data reduction computation.

The following question can be addressed by benchmarking: How is the local economy performing compared with a reference economy in a particular socioeconomic area? For example, one can look at employment growth, exports, ﬁrm creation, GDP, investment, and innovation.

How Is Benchmarking Used?
Originally created as a business development tool, benchmarking analysis is today widely used by local and national governments to assess competitiveness and formulate strategy. Benchmarking can be used to explain relative performance versus quantitative outcome measures, but it can also be used to compare qualitative factors and processes (such as how business support is provided). Benchmarking has become partic-

CASE STUDY
SAN FERNANDO (Philippines)
San Fernando, along with 65 other cities in the Philippines, is involved in the Asian Institute of Management’s (AIM) Philippines Cities Competitiveness Ranking Project (PCCRP), run by the Asian Institute of Management (AIM) Policy Center in collaboration with the United State Agency for International Development (USAID), the Asia Foundation, and GTZ. The benchmarking tool used for the project is based on the annual Institute of Management Development’s IMD World Competitiveness Yearbook and covers the following competitiveness issues: accessibility, cost competitiveness, dynamism of the local economy, human resources and training, infrastructure, linkages, responsiveness of local government, and quality of life. Data are derived from secondary sources, and a survey is undertaken specifically for the project. Scores for each indicator are converted into a 10-point scale based on global and national benchmarks. The output from the PCCRP is seen as extremely valuable in assessing local economic competitiveness. But it is difﬁcult to ensure that the project is fully institutionalised (or embedded in the ongoing local strategy process) so that updated data remain available. The League of Cities of the Philippines, which oversees the city development strategy (CDS) process in the country, is integrating the PCCRP survey output into the CDS process.

Edwin Huffman/World Bank Photo Library

Industrial facility in the Philippines

79

ularly popular as local economies increasingly design LED strategies using the competitive advantage framework. The main benefit of benchmarking is that it helps cities identify areas of improvement—a good way to measure performance of a local economy. Using this approach, it is often easier to see how well the local economy is performing when compared with an economy with similar growth conditions. Benchmarking analysis can be conducted for the overall economy, but is more often used to analyse speciﬁc sectors or topics. This is because benchmarking that relies heavily on quantitative statistical analysis can be resource intensive (with regard to acquiring data on the benchmark economy or economies in question). Cities typically benchmark aspects of their economy against the wider region and nation or against similar cities. So one of the biggest challenges to conducting this analysis is to ensure that the comparisons made are appropriate. The benchmark economy and the local economy should be

similar—based on the areas in which the economies compete and the economic structure. Using a substantially different benchmark economy in the analysis is not likely to give an accurate picture of regional performance or offer much guidance on what could be improved. The data used for benchmarking analysis can be acquired from several sources, including regional and/or national statistical bureaus in the benchmark economy. Or it may be possible to exchange information with the city or region chosen as a benchmark (e.g., directly with the benchmark city or working through a city network). In some cases, the benchmark economy may have conducted an investment climate or equivalent survey, and it may be a good idea to replicate that survey for comparison in the local economy.

What Key Inputs Are Required for Benchmarking?
Benchmarking analysis can be demanding in several ways. First, the analysis requires data for at least two cities: the city

ANALYSING THE DATA

Tips
Choose comparable data. It is important to make sure that the data being compared are strictly comparable, especially because of discrepancies in the measurements made by regional and national statistics bureaus. Choose reference economies. It is important to choose the reference economy to be compared with the survey city purposefully, not solely data availability. An important ﬁrst step is often to be clear about the purpose of the exercise. If a wide comparison for overall economic performance is the goal, choose a similar economy. But if there is a speciﬁc issue, market segment, or sector to be benchmarked, then speciﬁc competitors may need to be chosen. Contact benchmark economies. It is a good idea to contact government representatives and institutions in the reference economy chosen as the benchmark, because they may be willing to provide the data you needed for the survey (if, of course, the ﬁndings are shared with them). Use as a powerful motivator. Benchmarking can be a powerful communications and mobilisation device. And because it encourages stakeholders to look outside their own cities and provides hard data on relative strengths and weaknesses, cities may want to consider benchmarking in areas where they want to encourage changes in thought and action among stakeholders.

Further Information
For useful background information and perspectives on subnational benchmarking analysis, see BAK Basel Economics’ paper Regional Benchmarking and Policymaking at: http://bakbasel.ch/wDeutsch/bak/publications/papers/999 _regional_benchmarkingW3Dnavid W26149.shtml. For an example of the application of a benchmark analysis, see Glasgow Economic Analysis and Benchmarking study 2005 (beginning at page 14): http://www.glasgoweconomic facts.com/default-2.htm. For Web links to national statistical bureaus and comparative data worldwide, see: http://unstats.un.org/unsd/ methods/inter-natlinks/sd_natstat.htm. For the World Bank’s knowledge assessment methodology (benchmarking tool for assessing knowledge economy issues), see: http://web. worldbank.org/WBSITE/EXTERNAL/ WBI/WBIPROGRAMS/KFDLP/EXTUNIKAM/0,,menu PK:1414738~pagePK:64168427~piPK:64168435~the SitePK:1414721,00.html.

80

LED RESOURCE GUIDE

Geographical Information System (GIS) Mapping
What Issues Are Addressed by GIS Mapping?
The following questions can be addressed by GIS mapping:

studied and a reference economy. Getting access to these data and determining how the data were obtained can be challenging. Many cities obtain data for benchmark analysis through city networks, and others use data from established indexes (including competitiveness and quality of life indexes). Benchmarking does not require sophisticated econometric knowledge, but it does usually require basic statistical knowledge, in addition to some basic training in benchmarking. These skills are needed to ensure that the tool is applied effectively and that appropriate conclusions are drawn from the analysis. Speciﬁc software or analytical resources are not required. Overall, the tool has moderate resource intensity.

Where are ﬁrms located in the city or region? How does the sector composition of the economy play out spatially? Where are the concentrations of speciﬁc sectors? Where is industrial land located? Where are economic activities in relation to the labour force? Where are economic activities in relation to critical infrastructure?

How Is GIS Mapping Used?
GIS analysis is a computer-based tool that analyses spatial data from a database and displays the results in the form of

CASE STUDY
COPENHAGEN (Denmark)
Copenhagen used the GIS analysis as part of a wider analysis of the location patterns of their research-intensive industry. A key aim was to identify any industrial and spatial clustering effects. Ten research-intensive business sectors were identiﬁed (based on the expertise of the team of researchers), and a GIS analysis was performed on the database of 2,900 companies in 10 sectors (including information technology, electronics, energy, and transport). The analysis was conducted on each sector individually. First, a map was produced to show the location of the ﬁrms surveyed. Then a second map was produced to show the concentration of employment. The researchers for this study were looking for high concentrations of ﬁrms and

Sewer in a clothing factory, Copenhagen, Denmark

ha Top sen rand nn F Fi

m/

lfo Po

What Key Inputs Are Required for GIS Mapping?
Using this tool requires access to GIS software and training in how to use it. Some knowledge and understanding of spatial analytical techniques is also valuable. GIS does not require speciﬁc econometric knowledge, but this depends, in part, on the types of analysis to be conducted using this software. Using GIS tools requires fairly extensive data access. The type of data needed depends on the type of analysis to be

conducted. But it is always necessary to have detailed geographic information (including a database of maps that are linked to the latitude and longitude coordinates of speciﬁc locations). When the appropriate data and software are available, the analysis has a moderate resource intensity.

Courtesy of Chreod, Ltd. Canada

a map. By overlaying geographical information (actual geographic location of something on a map) and economic data (such as economic activity or employment data), GIS analysis can provide a powerful spatial analysis of the local economy. Historically, GIS analysis has been most often used by governments in land planning and natural resource management; but this tool is increasingly used by LED practitioners. For example, GIS is now used to map out ﬁrm distribution as part of conducting a cluster analysis, and to identify areas for upgrading and regeneration programmes. The example in ﬁgure 8.1 shows how GIS mapping was used in China to deﬁne the spatial patterns of economic growth as part of the Lanzhou Municipality CDS process.

F I G U R E 8 . 1 Spatial Patterns of Economic Growth, Lanzhou,China

81

ANALYSING THE DATA

employment in a radius of 3 to 5 kilometres (a parameter based on the expertise and judgement of the researchers). These steps were then complemented with an analysis of the factors behind location decisions for the ﬁrms studied (based on survey and qualitative interview data)

to/ T

opf oto .c

o.uk

Further Information
For a high-level discussion of GIS and its uses, see: http://www.library.yale.edu/MapColl/oldsite/gis/whatis.htm. For a step-by-step manual on the use of GIS software and related information, see: http://www.mapcruzin.com/learn_to_map/. For a comprehensive manual on establishing local GIS capacity, produced by the Massachusetts Geographic Information System, see: http://www.mass.gov/mgis/Getting_Started_With_GIS.pdf. For a manual for GIS developed by the United Nations Statistics Division, see: http://unstats.un.org/unsd/publication/SeriesF/SeriesF_ 79E.pdf.

LED RESOURCE GUIDE

Tips
Use pictures for impact. GIS analysis can be powerful way to make visual impressions of economic concepts that relate to geographic location and concentration. For example, nonexpert audiences can usually understand concepts illustrated by clusters—spatial concentrations of interconnected ﬁrms and institutions in related industries—shown on an actual map. Be cautious interpreting the results. It is important not to be misled by pictures derived from a GIS analysis. What appears to be an apparent industry concentration shown on a map does not always indicate some signiﬁcant underlying economic factor. Be sure to seek out alternative explanations. For example, is there really a cluster, or is the concentration shown on the map the result of local infrastructure or zoning regulations? Share the costs of GIS software and training. GIS tools are valuable in many aspects of city planning, including environmental planning, local economic development, and spatial planning. The costs for GIS software and training can be substantial, but sharing resources across departments may enable small cities to purchase this useful tool.

What aspects of the macroeconomic environment could impact the local economy? What are the implications of these factors for the local economy (now and in the future)?

82

How Is PEST/Trends Analysis Used?
Trends analysis looks at aspects of the macroenvironment (those largely outside of the control of local stakeholders) that are most likely to affect the local economy in the future. One of the most commonly used trends analysis tools is PEST, which looks at the political, economic, social, and technological factors of the external environment, as shown in ﬁgure 8.2. The PEST analysis (and trends analysis in general) typically focuses on qualitative issues and is designed to ensure that the competitiveness assessment and resulting strategies

F I G U R E 8 . 2 The PEST Trends Analysis Tool
POLITICAL

SOCIO-CULTURAL

PEST/Trends Analysis
What Issues Are Addressed by PEST/Trends Analysis?
The following questions can be addressed by PEST/trends analysis:

LOCAL ECONOMY

ENVIRONMENTAL

ECONOMIC

CASE STUDY
IVANO-FRANKIVSK (Ukraine)
The city of Ivano-Frankivsk used PEST analysis as a tool to generate inputs for their SWOT analysis. In particular, the PEST tool was used to identify external opportunities and threats that might affect the city in the short and medium term. Nine members from the Strategy Development Committee participated in a brainstorming session. The discussion was facilitated by technical staff from the United States Agency for International Development’s LED project and external consultants. The committee members shared ideas on potential future changes across the four areas speciﬁed by the PEST analysis, and then discussed whether, and how, these issues could present opportunities and/or threats for the city. For example, could the future reform of customs regulations and tax codes (legislative and political changes) present both threats and opportunities? Could reforms of this type reduce the city’s revenue base, but also simplify administration processes and increase transparency, consistency, and stability? In this setting, PEST analysis was seen as useful in highlighting potential drivers of change in Ivano-Frankivsk’s external environment. However, given the rapid changes being made in the city (and considerable political uncertainty in Ukraine at the time of the assessment in 2005), it was difﬁcult for the committee to determine how the city might respond to, or prepare for, possible external changes.

83

Tips
PEST analysis is not an alternative to SWOT analysis. PEST and SWOT analyses are often described as similar analytical tools, but they are not. PEST focuses on analysing the external environment, so it cannot be used on its own as a framework to assess the city’s overall competitiveness, as can be done with SWOT analysis. PEST is best used ﬁrst, as an input to the SWOT process. Make sure the analysis is dynamic. Analyses of the city’s external environment should aim to uncover key trends that are likely to affect the city in the future. The focus of a PEST assessment should be forwardlooking. Carefully analyse each factor. A simple list of factors that affect the local economy has limited value. To maximise the value of a PEST analysis exercise, it is important to think through what each of the factors means and decide which are likely to have the greatest impact on the local economy.

Further Information
For a brief discussion of and a free template for PEST analysis in a business context, see: http://www.businessballs.com/ pestanalysisfreetemplate.htm. For a discussion of the use of PEST analysis in a business context, see: http://www.themanager.org/Models/PEST_ Analysis.htm. For a set of guidelines to and a free worksheet for a PEST analysis, see: http://www.mindtools.com/pages/article/new TMC_09.htm.

take account of likely developments in the macroeconomic environment. A key advantage of PEST analysis is that it provides a structured and simple way to organise, analyse, and present a wide range of information.

Trends and PEST analyses lend well to participatory approaches. They are normally used as inputs to SWOT (strengths, weaknesses, opportunities, and threats) analysis—in particular, to add a more dynamic perspective to the analysis. Trends and PEST are also used to identify possible futures in scenario planning. (For a discussion of SWOT analysis, see chapter 9 of this Resource Guide.)

ANALYSING THE DATA

What Key Inputs Are Required for PEST/Trends Analysis?
Trends analysis can be effective with limited quantitative/ statistical data input, and it commonly draws on qualitative input from an internal and/or external expert panel. A PEST analysis can also be completed in a larger participatory forum with a trained facilitator. The analysis can be conducted with relatively limited resources and analytical capacity.

Sector Share Analysis
What Issues Are Addressed by Sector Share Analysis?
LED RESOURCE GUIDE

The following questions can be addressed by a sector share analysis: What is the basic structure of the city’s economy? Which sectors contribute the most to GDP, overall employment, and output?

insights into how global and national sectoral trends might affect the local economy. This type of analysis takes data on employment and/or output (such as GDP and total production) in each sector and calculates them as a share of the total economy. The results are then presented in a table or in a graph. Industry shares are usually based on widely-used standard classiﬁcation codes (such as SIC, SITC, NAICS, or HS). However, data are sometimes collected only at a broad level of aggregation (such as for primary, secondary, and tertiary industries). Because the sector share analysis only provides a static picture of the economy, it is most often used as a starting point for wider analysis. For example, it is often used with time series analysis, PEST/trend analysis, shift share analysis, input-output analysis, and location quotients. For cities that are conducting a local economy assessment for the ﬁrst time, sector share analysis is a critical starting tool that gives insight into the local economic structure.

84

How Is Sector Share Analysis Used?
Sector share analysis is probably the most common tool used to analyse the city economic structure. It identiﬁes signiﬁcant industries in the city-region and provides important

What Key Inputs Are Required for Sector Share Analysis?
The data required for sector share analysis are often available from local, regional, or national statistics bureaus. Cities that do not already have these data may want to use an industrial

CASE STUDY
KARU and BOBO-DIOULASSO
KARU (Nigeria) Karu used a sector share analysis in conducting its ﬁrst local economy assessment in 2002. City administrators in Karu knew very little about the basic composition of the local economy, so the sector share analysis was the single-most important tool used in the assessment. The analysis was designed to identify the main sectors of the city’s economy and, in particular, to assess the relative importance of the formal and informal sectors. Because of the near absence of formal statistics to describe the economy of Karu, data were collected using an industrial structure survey and an informal sector survey. The sector share analysis then compared the contribution of each of the sectors, by number of employees and the value of sales and output. Separate analyses were conducted for the formal and informal sectors, and these analyses were then compared to determine the relative importance of the different sectors. The use of this analytical tool was considered highly successful because it not only shed light on the basic structure of Karu’s economy, but it also helped to identify important stakeholders in the private sector. In particular, the analysis highlighted six main business sectors from which stakeholders were invited to participate in the city development strategy process. BOBO-DIOULASSO (Burkina Faso) Bobo-Dioulasso’s share analysis was derived from Burkina Faso’s national statistics. Much of the sectoral data were extracted from breaking down national aggregates at the local level to extract a rough overview of the sectoral structure of the economy. Field data were then used to corroborate the accuracy of the breakdowns of national data. Although there is a risk of inaccuracy, sector share analysis has the beneﬁt of generating signiﬁcant sectoral data that would otherwise be difﬁcult and costly to obtain.

Tips
Using modelling techniques to derive local sector structures. When locallevel data are not available on output by sector, it may not be necessary to conduct an industrial census or survey. When both the regional output data and city-level employment data are available, it is possible to model city-level output share by sector—if the relationships between employment and output in sectors are broadly similar at both levels of analysis. Although the results of modelling techniques are less precise, modelling can be a cost-effective approach and so is used by most cities around the world. Take the results further. As a stand-alone tool and a one-time snapshot of the industrial structure of the economy, sector share analysis does not identify which industries are growing fastest—or anything about the factors that underpin the sectoral composition. So this tool may be more effective when used along with qualitative information on the economy, acquired through interviews with industry experts and published reports, and other tools, such as benchmarking, time series, and PEST/ trend analysis. Try analysing aggregated sectors to understand broad trends. Combining a number of traditional sector deﬁnitions to create sector composites (for example, “knowledge sectors”) can be valuable to understanding how the local economy is structured with regard to broader global trends.

Further Information
For detailed information on how to use and construct tables and graphs, see: http://www.statsoft.com/textbook/stathome. html. For guidelines on creating pie charts and graphs, see South Africa’s Department of Provincial and Local Government guides to integrated development planning (Guide 4, Toolbox Part 2), at: http://www.thedplg.gov.za/subwebsites/ Publications_b.htm.
ANALYSING THE DATA

For an introduction to industry classiﬁcation and for guidelines on creating a snapshot of the economy, see: http:// cecd.aers.psu.edu/using_employment_data_to_better.htm. For guidelines on how to create a snapshot of the economy and an example of a sector share analysis for a U.S. county, see: http://www.economictoolbox.geog.psu.edu/ index.php. For a list of national statistical bureaus worldwide, see: http://www.oswego.edu/~economic/int-data.htm. For the Organisation for Economic Co-operation and Development’s (OECD) Structural and Demographic Business Statistics database (SDBS) with data on the sectoral composition for OECD member states, see: http://www. oecd.org/document/17/0,2340,en_2649_33715_36938705_ 1_1_1_1,00.html#SDBS. For data on the sectoral composition of non-OECD member countries, the International Yearbook of Industrial Statistics, published annually by UNIDO may be purchased at: http://www.unido.org/en/doc/3700.

85

structure survey to collect it. The following inputs are needed for this analysis: Data on the output (GDP or total production, for example) of the ﬁrms in the local economy by sector. Data on the number of employees of the ﬁrms in the local economy by sector. An agreed sector classiﬁcation system (in most cases cities should adopt the system currently used regionally or nationally). This analysis requires little or no external capacity or statistical or econometric knowledge. Sector share analysis does not require any particular resources other than data and human resources. When relevant local or regional data are available (and no additional data collection is required), the analysis has low resource intensity.

Value-Added Analysis
What Issues Are Addressed by Value-Added Analysis?
The following questions can be addressed by a value-added analysis: Which sectors and/or ﬁrms are the most important contributors to the local economy?

CASE STUDIES

´ POZNAN (Poland) and COPENHAGEN (Denmark)
Value added is often a preferred measure to other economic measures (including output and employment) in that it examines actual wealth creation. However, because access to value-added data at the regional level is often limited, mainly middle- and high-income cities use value-added analysis. In Copenhagen, however, total regional value added was divided by the average number of hours worked by employees in a year to serve as a proxy for labour productivity. This then fed into a benchmarking exercise in which the Copenhagen city-region was compared with six other European cities in four areas: innovation and specialisation, entrepreneurship, use of information and communication technologies, and human resources. The percent increase in labour productivity between two years (1995 and 2000) was estimated for each city in the benchmarking exercise and displayed in a histogram as a key measure of economic performance. The consultants conducting this analysis saw the great advantage of using the value-added measure (rather than GDP per capita, for example) because value added accounted for differences in average working hours in the benchmark economies

LED RESOURCE GUIDE

In Pozna´ , value-added analysis was used to demonstrate the diversity of n the local economy and highlight areas of strength (its services sector, for example). However, the analysis was only conducted at a high level of sectoral aggregation (agriculture, industry and construction, and services) because of the lack of data—a problem frequently encountered at the regional level.

86

How much of the output of local ﬁrms is produced within the city? In which sectors is our economy adding higher or lower value?

How Is Value-Added Analysis Used?
Value added is measured as the sales value of goods and services less the cost of inputs (materials, parts, and services) used to facilitate production. Alternatively it can be seen as the sum of payments made to labour by ﬁrms in the city plus payments made for these ﬁrms on investments. Value-added analysis provides several important indicators on the local economy. First, as gross value added (GVA), an alternative measure to gross domestic product (GDP), this tool is a fundamental indicator of the overall economy. Value added is seen as an important measure of ﬁrm contribution to the local economy, because it looks at actual wealth creation, not just employment or output. By taking value-added data of each ﬁrm and/or sector and calculating it as a share of the total economy, value-added analysis can also be a useful approach in conducting a sector share analysis. Moreover, dividing value added by one worker, one gets a commonly used measure of productivity.

The most common way of estimating value added is by using input-output tables, which essentially display the value of input and output (goods and services sold) by supplier and buyer respectively. (See section on input-output analysis for further information on this technique.) Another way to collect value-added data is by reviewing company accounts (see the further information box on page 87 for an example), which may be available at a government agency; if not, relevant data may need to be collected using an industrial structure survey tool.

Tips
On using estimated data. When there are no city-level data on value added, it is possible to estimate the data using the national inputoutput account. This is a less costly way of obtaining value-added data than through primary data collection; but the data are also likely to be less accurate. City-regions often are more dependent on external trade than their national economy overall, so estimates based on national data may underestimate the city-regions’ dependence on imports. More sophisticated modelling techniques supported by additional data can help improve accuracy.

Munich International Trade Fair Building

Further Information
The U.K. Department of Trade and Industry collects value-added data on 800 U.K. and 500 European ﬁrms through its Value Added Scoreboard. For more information, see: http://www.innovation.gov.uk/value_added/home.asp?p=home. For guidelines on creating a local economy “snapshot,” see: http://cecd.aers.psu.edu/using_employment_data_to_better.htm. For information on how to calculate value added based on the United Nations System of National Accounts, see: http://unstats.un. org/unsd/sna1993/tocLev8.asp?L1=16&L2=5. For an example of the use of value-added data in sector share analysis in the U.S. state of Oklahoma, see: http://pods.dasnr.okstate.edu/ docushare/dsweb/Get/Document-982/F-910web.pdf.
ANALYSING THE DATA

Many textbooks in regional science cover value added in relation to input-output tables. For information on how to read value added and to calculate GDP from input tables, see Regional Economics and Policy (3rd edition), by H. Armstrong and J. Taylor, published 2000 by Blackwell Publishing.

Value-added data at the regional or city level is often difﬁcult to obtain, and conducting an industrial census or industrial structure survey is often too resource intensive. Therefore, value-added data are typically modelled (based on regional and national data). Because the value-added analysis offers a simple snapshot of the contribution of ﬁrms to the economy, it is rarely used as a stand-alone tool. It is almost always used in conjunction with sector share analysis and time series analysis.

Economic Base Analysis
What Issues Are Addressed by Economic Base Analysis?
The following questions can be addressed by an economic base analysis: How much of the city’s economy is driven by meeting the local population needs versus selling products and services outside of the city (exports)? Which sectors and types of ﬁrms (by size) drive the city’s export economy? What proportion of the labour force in the city works in the export sectors?

Courtesy of Messe Munchen International

87

What Key Inputs Are Required for Value-Added Analysis?
Value-added analysis requires detailed data on ﬁrm accounts, including the value of sales and the cost of inputs. Since these data are not always available at the local or regional level, the data can be estimated using national input-output tables, although sometimes at the cost of accuracy (see tips box on page 86). Alternatively, an industrial structure survey can be used to collect local-level data on ﬁrm accounts. Although value-added analysis is relatively straightforward, it requires some level of econometric knowledge and comfort in reading input-output tables. However, it does not require any particular resources other than data and human resources. When comparable data are available at local and national levels (and no additional data collection is needed), the analysis has a low resource intensity.

How Is Economic Base Analysis Used?
Economic base analysis (also called export base analysis) is designed to analyse the broad economic structure of the local economy. It does this by dividing the economy into two sectors: 1) the basic or export sector (which includes all output—goods and services—sold outside the borders of the city or region), and 2) non-basic sector (which includes all output that is sold within the local economy, for example, output from local grocery retailing, hairdressing, restaurants, and other local services). Economic base analysis originates

CASE STUDY
MUNICH (Germany)
Munich used economic base analysis to forecast employment levels (for total employment and at a disaggregated sectoral and spatial level) for the city and its wider city-region for the years 2005–2015. Munich conducts employment forecasts every ﬁve to six years, in an effort to understand likely changes to economic structure and employment. Economic base analysis formed the backbone for the forecast model used in Munich. Econometric modelling was required since employment data are not captured by the ofﬁcial census in Germany and are not available at a spatially disaggregated level. The analysis divided the economy into three segments: the basic sector (exported products and services), the non-basic sector (local products and services), and the other services sector (which includes the public sector and other institutions with employment patterns that do not necessarily follow market rules). Local economic activity was assigned to this third sector based on the assumption technique, whereas economic activity was assigned to the basic and non-basic sector based on location quotient (measured by employment levels in Munich versus other metropolitan regions in Germany). Capturing growth trends in the basic sector was central to determining the employment forecasts for Munich. These growth forecasts were determined through a shift share analysis—in which the model ﬁrst examined the impact of development trends in the overall national economy and then considered potential regional variation above/below the national trend for the various industrial sectors in the basic sector. A simpler approach, based on regional population growth estimates, was used to determine the growth forecasts for the non-basic sector.

88 from economic base theory, which stipulates that inﬂow of money generated from the export sector is the main source of growth in an economy and determines the rate of employment and employment growth of the non-basic sectors serving local consumption. It is seen as a valuable tool, because it can be used to make relatively simple forecasts on income and employment for the local economy. Constructing an economic base analysis involves two main steps: 1. Determining the basic and non-basic sectors: There are two broad approaches to assigning local economic activity to basic and nonbasic sectors. One method, the assumption technique, assumes that certain sectors, such as mining and manufacturing, are wholly basic, while others, services, for example, are wholly non-basic. This is quick and easy for analytical purposes, but inaccurate— increasingly so as the services sector becomes more globalised. The second and more common method involves calculating location quotients for each sector (employment shares of a sector vis-à-vis a reference economy, normally the national economy) and assuming that any

LED RESOURCE GUIDE

Tips
Overestimating importance of exports. An important caveat of economic base analysis and its theoretical underpinnings is that it sees the exports as the sole driver of economic growth, neglecting the role of, for example, investments and productivity. When assessing the local economy using economic base analysis, it is important to recognise that there is considerable scope for growth outside of export; this is particularly true for larger cities and city-regions. Be careful comparing economic bases across locations. Economic base analysis is designed for understanding a city’s economic structure, mainly for conducting forecasting and assessing the possible impacts of possible future scenarios on the city (see scenario planning for further information). As such, it is generally not appropriate to compare basic versus non-basic structures or even base multipliers across cities. Be clear on city boundaries. The economic base analysis is focused on making the distinction between local and external economies. Therefore, deﬁning clearly the “local economy” is critical. Is the local economy a city? A city-region? What constitutes the wider external economy? Is it a region? A nation?

Further Information
For a discussion of the theory and application of economic base analysis, see: http://garnet.acns.fsu.edu/~tchapin/urp 5261/topics/econbase.htm. For guidelines and equations for economic base and multiplier models, see: http://www.rri.wvu.edu/WebBook/ Schaffer/chap02.html. For an example of the use of economic base analysis by the Canadian city of Dryden, see: http://www.dryden.ca/ website.nsf/DrydenSocioEconomicFinalReportFeb05.pdf? OpenFileResource#search=%22%22economic%20base%2 2%20dryden%22. For an accessible step-by-step guide to multiplier analysis, see: http://www2.sjsu.edu/faculty/watkins/EPM01.htm. Most textbooks in regional science or local economic development have good explanations of the role of economic base theory and analysis. See, for example: Local Economic Development: Analysis and Practice, by John P. Blair, published 1995 by Sage Publications.

Data on national and local-level employment by sector. See location quotient below for further information on data requirements for this approach. To analyse economic base on output or value added, corresponding local-level data on this are needed, or the local-level data will need to be modelled based on regional or national data. This analysis is moderately complex and requires capacity in econometric analysis. It does not require any particular resources other than data and human resources, although spreadsheet software or other statistical software (STATA, SPSS) would be useful. If comparable data at national and local levels are available and no additional data collection is required, the analysis has moderate resource intensity.

Location Quotient
What Issues Are Addressed by the Location Quotient?
The following questions can be addressed by the location quotient:

89

employment above the reference economy average is basic. 2. Calculating base multipliers: The base multiplier calculates the ratio of the total employment in the local economy to the basic employment in the economy: Base multiplier = total employment year x / basic employment year x The base multiplier gives an understanding of how changes in employment in the basic sector will inﬂuence the overall economy.
Note: Economic base analysis can be calculated on both output and income, but it is typically calculated based on employment data (often readily available at the local level).

How specialised is the city’s economy? In which sectors is the local economy specialised and concentrated?

How Is the Location Quotient Used?
Location quotient is one of the most widely used measures of specialisation and industrial concentration of a local economy. The location quotient takes the relative size of any sector (most commonly measured by employment or output) and compares it with equivalent rate in a reference economy (usually the national level). Therefore, this tool calculates how closely the local economy mirrors the structure of the national economy and in which sectors the local economy is more or less specialised. Typically, this is the primary tool used to identify where clusters may exist in the local economy (see cluster mapping, page 99, for further information). This tool is also often used to identify the import and export sectors in the local economy, as a part of economic

What Key Inputs Are Required for Economic Base Analysis?
The data required for this analysis may not be readily available from national and local statistical sources. For this analysis, the following are needed:

ANALYSING THE DATA

CASE STUDY
TORONTO (Canada)
The city of Toronto has structured its most recent economic development strategy in 2000 around clusters, so analysing location quotients of main sectors in the city economy was fundamental to developing its city strategy. The city ﬁrst identiﬁed the clusters that were most important to the Toronto economy in terms of employment. For these 10 clusters, an employment-based location quotient was then calculated (based on average employment concentrations for Canada and the United States combined). This allowed for an assessment of the degree to which Toronto could be considered specialised on a North American basis in priority clusters.
LED RESOURCE GUIDE

90

Toronto took the analysis further by combining this level of specialisation with benchmarks of relative growth rates and overall employment. This approach allowed for: 1) an assessment of Toronto’s relative competitiveness performance versus other cities in each sector; and 2) an assessment of the relative importance and competitiveness of each sector for the city. This analysis found that Toronto’s relative specialisation varied substantially among the 10 sectors, with biotech and business and professional services showing particularly high relative specialisation in the city.

Skydome and the CN Tower in Toronto, Canada

base analysis (with exports deﬁned as sales outside of the local economy, not necessarily outside of the country). Surplus employment or output in any sector of the local economy relative to the reference economy is assumed to be involved in exports, because it is in excess of what is needed to serve local demand. The location quotient is simply the ratio of the percentage of local employment or output in any sector to the equivalent percentage for the reference economy. It can be expressed as: LQ = ei /e Ei /E

score of greater than or smaller than 1 means that the local economy has a greater or smaller share of that sector than the reference economy, respectively. Therefore, by reviewing these ratios, the major import and export sectors in the local economy are easily identiﬁed. Locational Gini index, adjusted geographic concentration index, and entropy indexes and decompositions are similar, but more complex, spatial indexes used to measure industry concentration (see table 8.5). For information on an alternative measure of specialisation, see specialisation index.

What Key Inputs Are Required for the Location Quotient?
The data required for this analysis are often readily available from national and local statistical sources (especially employment data). The following data are needed for this type of analysis: Output and/or employment data for the city, by sector

where ei = local employment in industry i, e = total local employment, Ei = national (or reference economy) employment in industry i, and E = total national (or reference economy) employment. A ratio of 1 indicates that the local economy and the reference economy have an identical share of an industry; a

Mattes/Andia.fr/StillPictures

Tips
Get data that are progressively more detailed. The level of data aggregation matters—both in terms of how much work is involved and how much is gained from the analysis. It may be useful to start at a high level of aggregation (using two-digit standard industrial codes, for example) to identify the broad pattern of specialisation. For sectors where a specialisation or strong export sector is detected, location quotients can then be calculated at a ﬁner level of aggregation. This will help cities to identify speciﬁc subsectors that are driving the values seen at higher levels of aggregation. Be cautious of underestimates for exports. It is important to remember that the location quotient tends to underestimate the size of the export (basic) sector, because it assumes there is no cross-hauling (no simultaneous import and export of the same type of product in a region). Be sensitive to spatial scale of measurement. The score for the location quotient will depend on the spatial scale of measurement. This is particularly important to take into account when using regional employment data to measure specialisation at the city level. If the difference between the economic structure and the size of the region and the city is substantial, then the location quotient may give an incorrect impression of specialisation. Probe for implications. It is critical to examine the location quotient results and probe for the wider implications. Is the local economy highly dependent on particular sectors? Is this dependence a strength? Or, how might this dependence be problematic?

Output and/or employment data at the national or regional level, by sector Location quotient analysis is moderately complex and requires some minimal capacity in econometric analysis. Other than data and human resources, this type of analysis does not require particular resources, although spreadsheet software is useful. If comparable data at national and local lev-

els are available and additional data collection is not required, location quotient analysis has a moderate resource intensity.

91

Specialisation Index
What Issues Are Addressed by the Specialisation Index?
The following questions can be addressed by the specialisation index: How specialised or diversiﬁed is the local economy? To what extent is the local market dominated by a few ﬁrms or a few sectors? How reliant is the city on certain ﬁrms or sectors?

Further Information
For information on how to calculate the location quotient, see the Florida State University Web page at: http://garnet. acns.fsu.edu/ ~tchapin/urp5261/topics/econbase/lq.htm. For guidelines on how to calculate the location quotient, see: http://www.rri.wvu.edu/WebBook/Schaffer/chap02.html# Heading14. For an example of how location quotients are constructed and presented for U.S. counties, see: http://www.economic toolbox.geog. psu.edu/index.php.

How Are Specialisation Indexes Used?
Specialisation indexes measure the relative concentration or diversiﬁcation of a city’s economy. They can be used to understand concentration in, and reliance on, certain sectors and/or ﬁrms. Two examples of specialisation indexes follow: 1. Herﬁndahl-Hirschman Index (HHI). HHI is a measure of market concentration that indicates the extent to which the market in question is dominated by a few ﬁrms. The

ANALYSING THE DATA

CASE STUDY
RAFAELA (Argentina)
In Rafaela, the Herﬁndahl-Hirschman index is created in relation to the industrial census the city conducts every ﬁve to six years. (See industrial structure survey on page 62 for further information on the industrial census.) Rafaela sees the HHI as an important tool because it indicates the extent to which the local economy relies upon the performance of certain key ﬁrms. Ideally, Rafaela would like to compute the index for each sector. However, the small number of businesses in the city makes it difﬁcult to produce statistically robust results.
LED RESOURCE GUIDE

tool is complemented with a descriptive data analysis of the evolution of the market share for the top 4, 10, and 20 ﬁrms. The analysis also examines the evolution of sales, employment ﬁgures, and expansion rates by size of ﬁrm. In 2000, the HHI was calculated by estimating each local (surveyed) ﬁrm’s share of total sales, squaring it, and adding them all up. The resulting index of 1216 indicated a moderate degree of concentration.

To compensate for this shortcoming, the HHI is incorporated into a wider analysis of the composition of the local economy in Rafaela. Speciﬁcally, the

92

measure has traditionally been used by competition boards to supervise mergers and other structural changes in the private sector, but has also been used by cities to determine the extent to which the local economy relies upon the performance of a few ﬁrms. The formula for HHI is:

indexed (on a 0 to 100 index). (See growth indexes on page 75 for further information on this technique). See location quotient for other measures to calculate specialisation.

What Key Inputs Are Required for Specialisation Indexes?
For the purposes of a local economy assessment, si can be calculated as the output or (more likely) employment of a ﬁrm or sector i in the market, and n is the number of ﬁrms or sectors. So the employment share of each ﬁrm or sector is squared and the resulting numbers are added up. A resulting score between 1000 and 1800 indicates a moderate concentration; a score above 1800 indicates a high degree of concentration. 2. Tress index. A tress index measures the degree of concentration of a city’s economy on a sector basis. A tress index of employment contribution is determined by calculating each sector’s contribution to the local economy; multiplying each sector’s total employment by its appropriate weighting (that is, according to its share of the economy total); and calculating the sum totals of the weighted values for each sector. The totals will vary depending on the number of sectors in the economy and are then normally Computing the specialisation indexes is relatively straightforward and requires no speciﬁc econometric knowledge or

Tips
Complement with qualitative analysis. Quantitative data on ﬁrm and sector specialisation provide important insights into the structure of the local economy. However, in order to assess and understand the implications of apparent concentrations and specialisations, it is critical to also understand details on the activities and scope of these sectors and ﬁrms in the city, in addition to global trends. Be sensitive to spatial scale of measurement. The score for specialisation indexes will depend substantially on the spatial scale of measurement. If the difference between the economic structure and the size of the region and the city is substantial, then the specialisation indexes may give an incorrect impression of specialisation.

Further Information
For basic information and an online calculator for estimating the HHI, see: http://www.unclaw.com/chin/teaching/antitrust/herﬁndahl. htm. For information on how to calculate the HHI, see: http://www.usdoj.gov/atr/public/testimony/hhi.htm. For an overview of the Tress index and how to use it, see the Development Bank of Southern Africa’s (DBSA) Guidelines to Regional Analysis, at: http://www.dbsa.org/Research/Pages/Publications.aspx. For a critical discussion of entropy indexes as measures of industry concentration, see: http://faculty.smu.edu/maasoumi/Pdf%20Files/ MS2rev1.pdf.
ANALYSING THE DATA

The following are examples of specialisation indexes currently in use in local-level economic assessments:
● To see how the HHI is applied in Rafaela (Argentina), see Censo Industrial Rafaela 2000 under “publicaciones” at:

http://www.rafaela.gov.ar/es/Publicaciones-ampliar.aspx?p=20 (document in Spanish).
● For the application of the Tress index in the city of Tshwane (South Africa), see page 9 of the status quo analysis at:

http://www.tshwane.gov.za/idp2004.cfm.

93 software tools (with the exception of basic spreadsheet software). The analysis does, however, require detailed information on employment (and possibly output) for ﬁrms and sectors in the city economy. When the appropriate data are available, this tool has a relatively low resource intensity.

How Is Shift Share Analysis Used?
The shift share analysis assesses the performance of the sectors of a local economy—typically measured by employment growth—relative to a larger reference economy (most commonly the national economy). This tool is seen as a relatively simple and effective way to measure the competitiveness of both individual sectors and the overall economy. Speciﬁcally, the shift share analysis calculates how much of the employment growth experienced by a local economy in a speciﬁc time period can be accounted for by: 1) the economy’s mix of sectors, because different sectors grow at different rates; 2) the national growth rate, because a certain similarity between national and local employment growth is a reasonable expectation; and 3) local factors, because a local economy may possess a competitive advantage in certain sectors. If a city’s employment in a sector is growing at a faster rate than the national employment growth in the sector would suggest, the local economy is assumed to possess a competitive advantage in that sector. Calculating the shift term for a sector involves ﬁrst calculating the growth rates for the local economy and for the reference economy (either the regional or national economy).

Shift Share Analysis
What Issues Are Addressed by Shift Share Analysis?
The following questions can be addressed by a shift share analysis: How well are the different sectors in the local economy performing relative to a reference economy? How much of the employment growth experienced by a local economy can be accounted for by the national growth rate? How much of the employment growth experienced by a local economy can be accounted for by the mix of industries in a local economy? How much of the employment growth experienced by a local economy can be accounted for by local factors?

CASE STUDY
GLASGOW (Scotland)
Shift share analysis was used in Glasgow as part of a 2005 economic assessment study, Glasgow Economic Analysis and Benchmark Report, conducted by BAK Basel Economics. The shift share analysis was used speciﬁcally to understand the drivers of growth in the two most important sectors of the Glasgow economy: ﬁnance and business services, and the public sector. Speciﬁcally the analysis focused on disaggregating the regional growth rate in these two sectors into four main contribution effects (global, structural, national, and regional). Conducting an analysis at these four levels of disaggregation required substantial economic data as well as fairly sophisticated analytical techniques and expert resources. The results of the analysis were particularly useful for Glasgow in that they stripped out some of the structural issues that affected growth in the city and helped to isolate city-speciﬁc competitiveness issues, often masked in a more static type of analysis. For example, the shift share analysis in Glasgow found that although the city performed well overall in business services (as was the conventional wisdom), the regional effect was particularly strong for information technology services, slightly above average in banking, and particularly weak in real estate.

LED RESOURCE GUIDE

Port and city view of Glasgow, Scotland

94

This is calculated as:

Tips
Growth rate = (e2 – e1)/e1 where e2 = employment at time period 2, and e1 = employment in time period 1. The shift term is then calculated as: Growth rate sector x (local economy) – growth rate sector x (reference economy) If the shift term is positive, the local economy is growing faster than the reference economy in the speciﬁc sector. But a negative shift term indicates that the local economy is growing slower than the reference economy in the sector. It is important to note that a sector may experience a declining employment rate and simultaneously increasing output, so a local economy may actually possess a competitive advantage in that sector (despite shift share analysis results indicating otherwise). Therefore, the traditional employment-based shift share analysis has been extended to
Be sensitive to spatial scale of measurement. The outcome of the shift share analysis will depend on the spatial scale of measurement. This is particularly important to take into account when using regional employment data to measure specialisation at the city level. If the difference between the economic structure and the size of the region and the city is substantial, then the shift share analysis is likely to give an incorrect impression of specialisation. Complement the analysis with industry experts and literature. The shift share analysis only indicates which sectors of the local economy seem to possess a competitive advantage, not the source of this competitive advantage. By consulting industry experts and literature, it may be possible to identify the factor(s) underpinning a local competitive advantage—for example, quality of local endowments, availability of factors of production, and market access for particular products.

(FREELENS Pool) Werner/StillPictures

Further Information
For a step-by-step guide for shift share analysis developed by Penn State’s Centre for Economic and Community Development, see tool number 4, at: http://cecd.aers.psu.edu/using_employment_data_to_better.htm. For a calculator developed by the Department of Geography at the University of Shefﬁeld in the United Kingdom, for use in conducting shift share analysis, see: http://www.shef.ac.uk/geography/teaching/shift share/shift share.html. For an explanation and practical application of a shift share analysis for a U.S. county, see: http://www.economictoolbox.geog.psu.edu /index.php. For an in-depth and rigorous approach to shift share analysis, see Regional Impact Models by W. A. Schaffer at: http://www.rri. wvu.edu/WebBook/Schaffer/index.html. For a practical application of shift share analysis in the U.S. county of Lauderdale (Mississippi), see pp. 24 to 29, at: http://www. embdc.org/researchpublications.html. Most textbooks in regional science and/or local economic development generally cover the shift share analysis extensively. For a discussion and a practical application of the Total Factor Productivity approach to shift share analysis, see Regional Economic Development: Analysis and Planning Strategy, by R. J. Stimson, R. R. Stough, and B. H. Roberts, published 2002 by Springer-Verlag.
ANALYSING THE DATA

95 include average factor (capital and labour) productivity for sector and/or region. What are the economic linkages between sectors in the local economy? How do changes in one sector affect other sectors in the local economy? How do changes in sectors affect overall economic activity in the local economy?

What Key Inputs Are Required for Shift Share Analysis?
This type of analysis requires local and national (or regional) employment data by sector for at least two separate years. For the extended version of the shift share analysis, productivity data are also needed, overall or by factor (capital and labour) on a national (or regional) and local level. Other than data and human capacity, the only resource required to conduct shift share analysis is basic spreadsheet software. The analysis is of moderate complexity and requires some limited econometric capacity.

How Is Input-Output Analysis Used?
This tool is widely used in local economy analysis to determine linkages between sectors in the economy, by breaking down inputs into each sector (by contributing sector) and outputs from each producing sector (to consuming sectors). An input-output table therefore provides a summary of the transactions occurring within an economy over a selected time period, showing, for a given industry, the industries from which it purchases and the industries to which it sells. Input-output tables also show the use of industry production in private and government consumption, and the use in investment and sales outside of the region (exports). Table 8.4 provides a basic example of how a simple input-output table might look for a typical three-sector economy.

Input-Output Analysis
What Issues Are Addressed by Input-Output Analysis?
The following questions can be addressed by an input-output analysis:

CASE STUDY
BRISBANE (Australia)
The city of Brisbane used input-output analysis for projecting the regional economy in its 1999 and 2006 city strategies. For example, the 1999 strategy developed a detailed input-output transaction table that traced out the monetary value of transactions between industry sectors in the regional economy. The analysis provided both a concise, descriptive snapshot of the economy at that particular point in time and a matrix representation of the regional
LED RESOURCE GUIDE

economic structure. In particular, the analysis was used to demonstrate the nature and impact of transactions between industry sectors and to show how these transactions could change over time. Because of the complexity of the analysis, Brisbane used third-party specialist capacity to conduct the input-output analysis.

T A B L E 8 . 4 Example of an Input-Output Table for a Three-Sector Economy
Economic activities
Agriculture Manufacturing Services Labour

96

Inputs Inputs to Inputs Total to agri- manufacto Final outculture turing services demand put
5 10 20 20 15 40 80 20 5 20 100 40 20 50 50 0 45 120 250 80

These data need to be disaggregated by sector and available at the regional level. (Note: Where statistical data are not available, data can be collected through a detailed survey of all industry sectors in the local economy. However, this can be time and resource intensive.) Input-output analysis is complex and requires signiﬁcant capacity in econometric analysis.

Tips
The input-output analysis is an essential tool in local economic planning that can be used for a range of analyses. It is useful in understanding the structure of the local economy (especially the linkages between sectors and ﬁrms), and therefore it is frequently used in relation to value chain and cluster-mapping exercises. (See cluster mapping on page 99 and value chain analysis on page 101, for further information). This tool is also important for seeing how changes in one industry affect another. It is therefore of particular use in economic forecasting or scenario planning as part of the economic assessment and strategy process. (See scenario planning on page 118 for further information).
Exert caution with using estimated data. Input-output tables can be estimated using national input-output accounts. This is a less costly way of getting value-added data than through primary data collection, but the data may also be less accurate. Regions tend to be far more dependent on external trade than the national economy, so estimates based on national data may underestimate the dependence on imports Be aware of structural inter- and intra-industry changes. If the inputoutput links between sectors and ﬁrms are changing rapidly, then the role of input-output analysis in forecasting is quickly made redundant. In cases where the local economy is believed to be undergoing signiﬁcant structural change, it may be best to consider collecting additional information about these changes through surveys to account for them in a forecasting exercise.

What Key Inputs Are Required for Input-Output Analysis?
This type of analysis typically requires detailed statistical information on sectors in the local or regional economy.

Further Information
For a guide on constructing regional input-output tables (including examples) refer to the Web site of The National Institute of Economic and Industry Research in Australia: http://www.nieir.com.au/index.php?option=com.content&task=view&is=28&Itemid=114. For a discussion of the theoretical underpinnings and application of input-output analysis, see chapter 11 in An Introduction to Regional Economics, by E. M. Hoover and F. Giarratani, at: http://www.rri.wvu.edu/WebBook/Giarratani/chaptereleven.htm. For an in-depth discussion of the construction and application of input-output tables, see: http://www.rri.wvu.edu/WebBook/Schaffer/ index.html. For a discussion of the application of input-output tables in cluster mapping, see: http://www.rri.wvu.edu/WebBook/BergmanFeser/chapter3.htm#3.3.3. Most textbooks in regional science and/or local economic development generally provide extensive discussion of the input-output analysis. See, for example, Local Economic Development: Analysis and Practice, by John P. Blair, published 1995 by Sage Publications. For an overview of current issues in input-output analysis, see the International Input-Output Association (IIOA) Web site at: http:// www.iioa.org/.
ANALYSING THE DATA

Social Accounting Matrix
What issues Are Addressed by the Social Accounting Matrix?
The following questions can be addressed by the social accounting matrix: What are the economic linkages between sectors in the local economy? How do changes in one sector affect other sectors in the local economy?

How do changes in sectors affect overall economic activity in the local economy?

97

How Is the Social Accounting Matrix Used?
A social accounting matrix (SAM) is a general equilibrium model of the economy based on the principles of input-output analysis—essentially, it is an extension of the input-output model. The SAM uses a system of accounts framework to track economic ﬂows between the supply side and institu-

CASE STUDY
BOBO-DIOULASSO (Burkina Faso)
A social accounting matrix (SAM) was extensively used in Bobo-Dioulasso during the city development strategy (CDS) process, which was guided by the OECD and MDP ECOLOC process. The Bobo-Dioulasso SAM is a simpliﬁed input-output table, showing the economic interaction among sectors in the area. This approach was particularly well-suited to Bobo-Dioulasso because of the low level of statistical complexity required and because of this tool’s capacity to accommodate information on different socioeconomic aspects stemming from the surveys conducted within the CDS framework. The SAM was used to uncover accounting transactions and linkages in the city and its surrounding area and was a useful tool in identifying economic activity complexes in Bobo-Dioulasso. However, a drawback of using the SAM was that, even though it was an excellent method for identifying economic interaction, it was less effective at uncovering the more social and structural linkages.

Tips
Exert caution with using estimated data. SAM tables can be estimated using national input-output accounts. This is a less costly way to obtain value-added data than through primary data collection, but the data may also be less accurate. City-regions are often far more dependent on external trade than the national economy, so estimates based on national data may underestimate the dependence on imports. Be aware of structural inter- and intra-industry changes. If the linkages between sectors, ﬁrms, and households change rapidly, the role of SAM analysis in forecasting quickly becomes redundant. In cases where the local economy is believed to be undergoing signiﬁcant structural change, it may be best to consider collecting additional information about these changes through surveys to account for them in a forecasting exercise.

LED RESOURCE GUIDE

duction sectors (agriculture, manufacturing, and so forth) pay factors of production (labour and capital) for services rendered. The factors pass this money along to different types of households (rural households versus urban households). The households, in turn, pay the production sectors for purchases of food, clothing, and so forth. One beneﬁt of using a SAM is that it allows for substantial ﬂexibility in deﬁning the data to be analysed. It is possible to adopt a wide range of units of measurement and alternative deﬁnitions, while maintaining internal consistency in the model (that is, inputs must equal outputs). For example, it is possible to disaggregate the model at whatever level of detail is desired to explore speciﬁc issues such as links between growth, income allocation, and poverty.

What Key Inputs Are Required for SAM? tional accounts. It therefore helps to illustrate how income is derived from production activities and how it is distributed to the various socioeconomic groups in the economy. A basic SAM model contains three institutions: factors of production, household types, and production sectors. ProAnalysis using a SAM typically requires detailed statistical information on the local economy, as well as on government and household spending. The data should be available at least at the regional level to allow for effective local-level analysis. (Note: Where statistical data are not available, data can be collected through a detailed survey of all industry sec-

98

Further Information
For an introduction to using SAM at a subnational level, see: http://www.implan.com/library/documents/elements_of_the_implan_sam. pdf. For a paper describing the methods and data sources for constructing SAMs for small cities, see: Schwarm, Walter, and Harvey Cutler. 2005. “Multiple Labor Groups and Their Effects on Small City and Town SAMs and CGE Models.” Review of Urban and Regional Development Studies 17(2): 162–176 (downloadable via Blackwell-Synergy: http://www.blackwell-synergy.com/doi/abs/10.1111/1467940X.00069). For a general reference on SAM, see: Graham Pyatt and Jeffrey Round. 1985. “Social Accounting Matrices: A Basis for Planning,” at: http://www.worldbank.org/reference/ (type in author name in “Dpcuments and Reports” ﬁeld). For a detailed discussion on the construction of SAM in South Africa and comparisons with 11 other countries, see the Statistics South Africa discussion paper at: http://www.statssa.gov.za/Publications/DiscussSAM/DiscussSAM.pdf. For an example of the use of SAM as a regional economic model in Alaskan Fisheries, see: http://www.st.nmfs.gov/st5/documents/ Review_of_Regional_Economic_Models_in_Fisheries.pdf.

tors in the local economy. However, this can be time and resource intensive. There are also a variety of statistical approaches to model local-level data.) SAM analysis is complex and requires signiﬁcant capacity in econometric analysis.

To what extent are the ﬁrms and institutions in a cluster geographically concentrated?

How Is Cluster Mapping Used?
Cluster mapping (also referred to as cluster analysis) identiﬁes groups of tightly linked ﬁrms in related industries in a local economy. Popularised by management theorist Michael Porter, it has become one of the most popular tools in regional planning since the 1990s. A cluster generally refers to a group of ﬁrms and supporting institutions, such as universities and research centres, that operate in related sectors, are interlinked through trade and knowledge exchange, and operate in proximity to each other. Identifying clusters in a local economy generally involves two types of analysis: 1) assessing the degree of geographical

Cluster Mapping
What Issues Are Addressed by Cluster Mapping?
The following questions can be addressed by cluster mapping: What are the main clusters of economic activity in the local economy? What are the nature and strength of links between the ﬁrms and supporting institutions in a cluster?

CASE STUDY
MEDELLÍN (Colombia)
As part of a wider analysis of seven priority clusters in Medellín, the city’s Chamber of Commerce carried out a cluster mapping exercise. The results of this exercise form an important input into the strategic development plan for Medellín to be ﬁnalised in 2007. The seven clusters identiﬁed through the cluster mapping exercise were named as priority industries in the ﬁnal strategy. The aim of the cluster mapping exercise was to identify clusters with the greatest actual and/or potential contribution to economic growth in the city. The city ﬁrst identiﬁed seven industries (based on data on regional value added, exports, output, and employment) that fulﬁlled the above criteria. This analysis was carried out at a disaggregated data level, using the International Standard Industrial Classiﬁcation (ISIC) of all economic activities. Then all sectors and research institutions related to each of these seven sectors were identiﬁed based on: 1) a review of existing secondary sources, 2) statistical analysis, and 3) a series of workshops with industry experts and representatives from the largest ﬁrms in the seven sectors. Once the clusters had been identiﬁed, their competitiveness was assessed using Porter’s Diamond method (see the competitive advantage framework in chapter 9 of this Resource Guide for further information on this approach).

99

Construction work, Colombia

Edwin Huffman/World Bank Photo Library

ANALYSING THE DATA

Tips
Interpret the results of quantitative analysis with caution. A high location quotient for an industry sector does not necessarily indicate the existence of a cluster. Cities are likely to have high concentrations of some activities (services, for example) whether or not there is any actual clustering. Similarly, a low location quotient does not necessarily preclude the existence of a cluster in the region. And although a GIS analysis may suggest the possible existence of clusters, it is important to explore other possible explanations. Clusters are deﬁned not just by a physical concentration of ﬁrms, but also by their activities and interrelationships.
LED RESOURCE GUIDE

using location quotients or other indices of specialisation. A complement or alternative is to plot the location of ﬁrms and supporting institutions in geographical space using GIS mapping. The linkages between ﬁrms can be quantiﬁed using inputoutput tables—displaying the value of input and output (goods and services sold) by supplier and buyer, respectively. Additionally, linkages can be assessed through a more qualitative and participatory mapping exercise of the relationships among key ﬁrms and related institutions in the cluster based on input from workshops, focus groups, or interviews.

100

Complement quantitative data with qualitative insights. Quantitative tools such as input-output analysis can help to determine the degree of trade linkages between sectors in a city or region, but a cluster mapping exercise that focuses exclusively on quantitative tools may omit important qualitative aspects of a cluster. For example, inputoutput analysis does not account for the role of supporting institutions (such as higher education institutions) or the potential importance of nontrade linkages between ﬁrms (knowledge sharing, for example).

What Key Inputs Are Required for Cluster Mapping?
A highly quantitative cluster mapping exercise is data intensive and may involve data that are not readily available at city level. Overall, a cluster mapping analysis requires: Qualitative or quantitative information on the ﬁrms operating in the local economy, such as the interﬁrm ﬂow of goods and know-how, employment by ﬁrm and/or activity; and Data on output and employment for ﬁrms in the local and reference (most national or regional) economy, if calculating clusters based on location quotient. This analysis is moderately complex and requires capacity in econometric analysis as well as an understanding of

concentration of the group or sectors of ﬁrms; and 2) assessing the strength and nature of interﬁrm linkages. There is no standardised methodology for this analysis, but it most often involves both quantitative and qualitative techniques. The degree of geographical concentration is typically calculated

Further Information
For a useful introduction to methods and concepts relevant for cluster analysis and mapping, see: Industrial and Regional Clusters: Concepts and Comparative Applications, by E. M. Bergman and E. J. Feser, at: http://www.rri.wvu.edu/WebBook/Bergman-Feser/contents. htm. For information on the UNIDO cluster development programme that focuses on developing countries, see: http://www.unido.org/ doc/4297. For examples of cluster processes in developing countries, see: the Global Cluster Initiative Survey (GCIS) on cluster initiatives in Developing and Transition Economies, at: http://www.cluster-research.org/devtra.htm. Cluster mapping is not typically addressed in regional science and/or local economic development textbooks. R. J. Stimson, R. R. Stough, and B. H. Roberts dedicated a chapter to the various methods used in cluster mapping in Regional Economic Development: Analysis and Planning Strategy.

cluster theory and theories of competitive advantage. If the cluster analysis includes spatial mapping, GIS data and technology are required. Cluster mapping has a moderate to high resource intensity.

What are the obstacles for local ﬁrms in moving up the value chain?

How Is Value Chain Analysis Used?
A value chain describes a series of stages that create and build value in products and services. Value chain analysis, developed by management theorist Michael Porter, has most commonly been used in the private sector as a tool to build competitive advantage by identifying opportunities to build process efﬁciencies (by cutting costs, for example) or ﬁnding sources of differentiation. The process is now increasingly used as a tool in the context of local economy assessments, particularly in terms of working with key sectors in the local economy to help ﬁrms understand their current position in global value chains and identify obstacles and opportunities for upgrading to more lucrative parts of the chain. Value chain analysis slices up the production chain into the different activities (logistics, sales, marketing, production, research and development) and looks at the spread of ﬁrms in the different activities, the relationships between them, and the extent to which the ﬁrms in the different activities control the value chain. Figure 8.3 shows a basic value chain within a regional development model.

Value Chain Analysis
What Issues Are Addressed by Value Chain Analysis?
The following questions can be addressed by a value chain analysis: How are the different activities involved in production and distribution structured in the local economy (overall and within speciﬁc sectors)? ● How and to what extent are the ﬁrms in the value chain interlinked in terms of ﬂows of goods and services? ● Where in the value chain do the ﬁrms in the local economy sit? ● In what kind of activities are the ﬁrms in the local economy involved? What areas need improving to make local ﬁrms more competitive (product quality, innovation, reliability of supply and logistics, and so forth)?

101

CASE STUDY
DURBAN (South Africa)
The city of Durban puts substantial emphasis on identifying and supporting key sectors as part of the local economic development strategy. Durban makes regular use of value chain mapping in order to assess the dynamic interaction between local ﬁrms and between groups of local ﬁrms and national and global value chains. Speciﬁcally, Durban has found value chain assessment useful for two elements of the assessment. First, it helps them to understand the degree to which local ﬁrms are operating in lower value and marginal activities or are moving into high value and more complex activities. Second, it allows them to understand the active, or in some cases missing, linkages among local ﬁrms. Because the value chain tool requires substantial depth of data gathering and analysis, it is most often used to analyse a speciﬁc set of target sectors rather than for the economy overall. Typically, the analysis involves both quantitative and qualitative analysis and has been used successfully as part of participatory assessment techniques with the private sector. Durban ﬁrst used value chain analysis as part of a major economic strategy. The study, Durban at the Crossroads, mapped out Durban’s priority clusters and conducted value chain analysis of the most important of these. It helped to identify the strengths and weaknesses of Durban’s clusters and to establish the interventions required to ensure sustainable competitiveness for the city’s industries.

ANALYSING THE DATA

F I G U R E 8 . 3 A Basic Value Chain Within a Regional Development Model
THE ENABLING ENVIRONMENT RESEARCH/ INNOVATION ORGANISATION

VALUE CHAIN PRODUCTS and PRODUCTION PACKAGING LOGISTICS MARKETING SERVICES

COMMUNICATION AND INFORMATION TRAINING AND DEVELOPMENT LED RESOURCE GUIDE

SALES

MARKETS

FINANCING

Source: Kaiser Associates Economic Development.

102

Tips
Use local resources and industry experts. Value chain analysis will in many cases involve participation from the private sector. Local business groups, in particular, may possess considerable qualitative and quantitative data on the value chain in question and knowledge and expertise on the topic. They may have suggestions and contact details for ﬁrm representatives and experts to include in a participatory exercise. Understand enabling factors. Although understanding linkages among ﬁrms in the value chain is the main focus, it is also critical to identify how ﬁrm and sector performance is supported (or hindered) by the wider enabling environment. This includes factors such as customs and transport policies and business support services, which must be aligned to improve sector performance. Make sure the analysis is appropriate for the purpose. Although the value chain tool is useful, it is important to recognise that value chains differ signiﬁcantly from one industry to the next. As a result, value chain analysis is unlikely to be a relevant tool for analysing the local economy in its entirety, but rather only for analysing individual sectors of the economy.

There is no standardised methodology for conducting a value chain analysis. Instead, it often uses a range of quantitative and qualitative tools. Input-output analysis can be used to diagram and quantify the relationships between ﬁrms in an industry value chain. A more qualitative and participatory mapping exercise of the relationships among ﬁrms in each value chain, as well as the degree of power of the ﬁrms (in different activities represented in the chain), can be used in addition to (or instead of) input-output analysis. The value chain analysis is also often integrated into cluster mapping exercises. (For more information on cluster mapping, see page 99.)

What Key Inputs Are Required for Value Chain Analysis?
The qualitative data required for value chain analysis are typically readily available; in most cases capturing this information requires facilitating a participatory process with the industry sector in question. However, rigorous quantitative analysis is data intensive and data may not be readily available at city level. Overall, value chain analysis requires: Qualitative or quantitative information on the ﬁrms in the value chain and the linkages between them, for exam-

Zhongshan Bridge over Yellow River in Lanzhou, China

Further Information
For a discussion of value chain analysis and theory in Value Chain Analysis for Policy-Makers and Practitioners by Hubert Schmitz, see: http://www.ilo.org/dyn/empent/docs/F204969253/VCA_book_ﬁnal.pdf. For a methodological and theoretical discussion of the value chain analysis, see “A Handbook for Value Chain Research” at: http:// www.ids.ac.uk/ids/global/manuals&handbooks.html. To view documents on value chain theory and analysis posted on the Institute of Development Studies Web site, see: http://www.ids. ac.uk/ids/global/valchn.html. For the “Participatory Value Chain Analysis” toolkit developed by the Enterprise Development Impact Assessment Information Service, see: http://www.enterprise-impact.org.uk/informationresources/toolbox/valuechainsanalysis.shtml.
ANALYSING THE DATA

ple, ﬂow of goods and services, employment by ﬁrm and/or activity, regional imports and exports by sector, added value by activity. An understanding of value chains theories and the role of value chains in deﬁning regional and ﬁrm competitiveness. This analysis is moderately complex and requires capacity in econometric analysis if it is based on quantitative analysis. In any case, it will require some understanding of theories of competitive advantage, of basic business strategy, and of the speciﬁc industries in question. This analysis does not require any particular resources other than data and human resources. If data at the local level

are available and additional data collection is not required, the analysis has a moderate resource intensity.

Panorama Images/The Image Works

Asset Mapping
What Issues Are Addressed by Asset Mapping?
The following questions can be addressed by asset mapping: What are the assets that might make the city more competitive? What are critical tangible assets? What are critical intangible assets?

103

CASE STUDY
ADEN (Yemen)
Asset mapping was an important analytical tool in the city development strategy process in Aden. In 2002, Aden’s city strategy process was built around taking advantage of what were seen as the city’s key infrastructure assets. The CORE, as they were deﬁned in Aden’s CDS process, included: The Port of Aden, The Aden Free Zone, and Aden International Airport. One component of the local economy assessment in this city involved undertaking a speciﬁc competitiveness assessment of the CORE, with regards to competing transport hub locations in the wider urban region. More broadly, Aden undertook a basic asset mapping exercise within the competitiveness assessment stage. As part of the Economic and Business Opinion Survey that was undertaken in the city during 2002, respondents were asked to name what they viewed as the city’s three most important assets (tangible or intangible). These assets were then further discussed and prioritised at a subsequent competitiveness seminar.

LED RESOURCE GUIDE

Coastal view of Aden, Yemen

How Is Asset Mapping Used?
Asset mapping is a relatively quick and simple tool for understanding local assets and how they can facilitate local economic competitiveness. The tool is used to document the main tangible assets (physical infrastructure and buildings such as transport, property, utilities, cultural amenities, and so forth) and intangible assets (knowledge, trust, cooperation, and so forth) available to the local economy. In the economic development context, it is most often applied at the community or small-city level.

104

Tips
Use visual aids. It is often a good idea to visualise assets, problems, and planning ideas by using photos and other visual tools. This is particularly important in processes with wide stakeholder participation, where not all participants may have sufﬁcient literacy skills. Be cautious in interpreting the results. Highly participatory processes such as asset-mapping are sometimes overly optimistic (or pessimistic) in interpreting sources of strength and weakness in the local economy. Information on local assets should be taken as a starting point for analysis and should be complemented with comparative analysis (with other cities and regions) and other objective analytical tools.

Asset mapping is often conducted by using a participatory approach in which local stakeholders help identify, rank, and prioritise the different local assets. In some cases, it also involves assessing factors that may threaten and/or strengthen these assets and developing strategies to support and enhance them. The type and number of participants will depend on the purpose of the exercise. If the tool is used as an exclusively analytical exercise, it may be best to choose participants based on their expertise. However, if the aim of the asset mapping exercise is also to sensitise stakeholders to local economic development and strategic planning, then it may be best to include a broader stakeholder group. Asset mapping can also be conducted through a more formal survey (a household survey or investment climate survey), but typically, a survey would only supplement a more participatory process. A main strength of asset mapping is that it generally supports a positive outlook as it focuses on opportunities rather than problems (which is often a drawback with participatory issues identiﬁcation and other participatory approaches). Therefore, it can be a more unifying and motivating tool for engaging stakeholders.

What Key Inputs Are Required for Asset Mapping?
The data required for this analysis are readily available from participants in the local economy. Asset mapping is a relatively

Ahmed Eiweida

Further Information
To view the step-by-step guide for asset mapping developed by The Canadian Rural Partnership, see: http://www.rural.gc.ca/conference/ documents/mapping_e.phtml. To view the step-by step guide for asset mapping developed by Community Builder, see: http://www.communitybuilders.nsw.gov.au/ download/Making_Headway_ToolKit.pdf. For a discussion of an asset mapping exercise with a focus on local residents, organisations, and community leaders, see: http://faculty. salisbury.edu/~mdforaker/IDIS/Mapping%20assets%20of%20community.pdf.

easy tool to use and does not require any speciﬁc econometric or analytical expertise. Using the tool effectively does require an experienced and appropriately trained facilitator, if using a participatory approach. Other than access to facilities to host a participatory session, the tool does not require any other speciﬁc resources. Overall, the tool has a low resource intensity.

● ●

What impact does this have on local ﬁrms? What impact does this have on the local economy? ● Is this a barrier to employment growth? What are future skill requirements? What are future trends in workforce skills? Is the city’s skills base competitive? Does its skills base match the strategic vision for the city?

105

Skills Audit
What Issues Are Addressed by a Skills Audit?
The following questions can be addressed by a skills audit: To what extent are the skill requirements of industry in the city covered? Are there any skill gaps or shortages in the local economy?

How Is a Skills Audit Used?
A skills audit is used to document the skills base and mapping against the identiﬁed needs of the local economy. This tool is typically used to: 1) determine current and/or future skill requirements of the local economy (overall or for speciﬁc industry sectors), and 2) determine the skill base needed to fulﬁl the city’s strategic vision. There are several possible ways to conduct a skills audit, including participatory assessments and more quantitative approaches that differ in methodology and the degree of expertise required. A ﬁrst step is often to review the local economic structure and identify any immediate changes to local employment (such as a planned opening or closing by a major local employer). This is of interest because the sectoral composition can indicate what skills are needed. For example, rapid expansion in the construction sector could indicate a need for more engineers and manual labourers, whereas growth in business services could indicate a need for more lawyers and accountants. A quantitative skills audit could simply involve using a ﬁrm survey to collect data on the perceptions of ﬁrms’ man-

Cranes in the port of Aden, Yemen

Tomas Sennett/World Bank Photo Library

ANALYSING THE DATA

CASE STUDY
TORONTO (Canada)
The most recent Economic Development Strategy conducted in Toronto in 2000 places substantial emphasis on competing in a knowledge-intensive economy through the human capital of its residents. To further assess its human capital competitiveness and needs and to develop a strategy for the future, the city undertook a detailed skills audit and skills mapping assessment, which culminated in the 2003 Toronto Labour Force Readiness Plan. Toronto made use of the Canadian Occupational Projection System (COPS) to assess the future surplus/deﬁcits on occupational requirements based on forecasted economic and sectoral growth. Against this demand-side assessment, Toronto assessed the current skills and education base of its workers using secondary data and an employer’s survey. The telephone-based survey covered 1,000 ﬁrms in greater Toronto, and captured both data and perceptions on existing labour skills, job requirements, and strengths and weaknesses in the city’s labour pool. What was seen as particularly valuable was combining data and information about the existing workforce in the city with forward-looking demandside assessments, in order to be in a position to understand the positioning of the city for future competitiveness. For more details on the methodology used and the development of Toronto’s Labour Force Readiness Plan, see: http://www.toronto.ca/ business_publications/labour_force_readiness_plan.htm.

106

LED RESOURCE GUIDE

agement (on the current and future skills demand) and to compare these data with the actual skill and education level of both the economically active population and students entering higher education. However, it could also involve constructing and analysing a matrix in which the number of ﬁlled posts and vacancies in terms of sector (usually standard industrial code [SIC] or harmonised system [HS] code classiﬁcation) and occupation (generally according to a national standard occupational classiﬁcation system). This is then typically complemented with an analysis of the supply of skills (for example, the level of education of the economically active population). To plan for future skills shortages, forecasting or scenario planning methods are often used along with a skills audit to identify future skills needs (see scenario planning in chapter 9 of this Resource Guide for more information on these tools).
Note: If the aim of the analysis is to map skills needs in relation to reaching a strategic vision, this may entail comparison with a reference economy (by using benchmarking analysis). For example, if a city wants to establish itself as a location for research and development ﬁrms in information technology, it can be useful to benchmark the city’s skill base against a city that has already established itself as a location for research and development ﬁrms.

Tips
Use local resources. To avoid duplication, possibly reduce costs, and take advantage of existing expertise, it is wise to make use of local business networks and educational institutions. These groups sometimes have substantial internal analytical capacity and data, so they may be able to help in the dissemination of results and provide information on future economic events that affect the availability of skills (such as the opening or closure of a local factory). Assess both supply and demand. A skills audit of the local economy requires a strong understanding of the supply side (existing skills in the local economy and those being developed for the future) and the demand side (how employers evaluate the local economy’s skills base and what skills they need for the future). It is important that the local economy analysis balances the two. Mix data and participatory inputs. Understanding the existing skills base and strengths and weaknesses of a local economy requires objective data and qualitative inputs from local stakeholders. A lack of data increases the risk of making assessments based on perception only; a lack of qualitative input from local stakeholders increases the risk of not understanding the nature and scope of issues relevant to effective strategic planning.

CDS Workshop in Lanzhou, China

Further Information
For a more in-depth discussion of skills audit, see: http://www.eprc.strath.ac.uk/eprc/Documents/PDF_ﬁles/R28TrainNeedsAssessin Scot.pdf. For one possible approach to a participatory skills audit, see: http://unauthorised.org/ronni/comdev/skillsaudit.html. For an example of a comprehensive skills audit that includes the use of scenario planning, see: http://www.yorkshirefutures.net/siteassets/ documents/YorkshireFutures/6/6/6645A51B-EB50-475E-ADFE-A6F9AB7E5DE3/FRESA%20labour%20market%20analysis.pdf. For an introduction to the U.K. system of Standard Occupational Classiﬁcation, see: http://www.statistics.gov.uk/methods_quality/ ns_sec/soc2000.asp.
ANALYSING THE DATA

A qualitative approach will generally take the form of a participatory workshop (see participatory issues identiﬁcation in chapter 7 for more information). In these cases, stakeholders, typically from the private sector and educational institutions but often including community organisations and skills experts, identify skills requirements and compare them with existing skills.

Stakeholder Analysis / Institutional Mapping
What Issues Are Addressed by a Stakeholder Analysis?
The following questions can be addressed by a stakeholder analysis: Who are the key stakeholders in the local economy? What are the speciﬁc interests and roles of these stakeholders? What is their inﬂuence on (and interests in) the strategic planning process? Who are the most important stakeholders to involve in the local economy assessment and/or strategy development process?

Courtesy of Chreod, Ltd., Canada

107

What Key Inputs Are Required for a Skills Audit?
The data requirements for a skills audit largely depend on whether a participatory or quantitative analysis approach is used. A minimum data requirement typically includes educational attainment levels of the population, number of students by career, and some type of data on the demand for skills by industry (qualitative or quantitative). For a more complex analysis, time series data on the above data on the sectoral composition of the economy, and data on the vacancies and jobs ﬁlled by occupation and sector, are needed. Data on future and current skill demand can be collected using an industrial structure survey or a business outlook survey. Similarly, the requirements for analytical capacity will vary. The most basic analysis is relatively simple, whereas the more complex analyses involving forecasting and scenario planning require substantially more analytical capacity and econometric skills. A trained facilitator is required for the participatory approach to conducting a skills audit.

How Is Stakeholder Analysis Used?
In stakeholder analysis, also called institutional mapping, important stakeholders are identiﬁed, and the relationships among stakeholders (along with their interests and inﬂuences in the local economy) are analysed. A stakeholder analysis can be conducted before the stakeholder consultations begin as part of a local economy assessment or strategy development process. The analysis can also be a valuable way to identify potential roadblocks and/or catalysts for moving forward on a chosen strategy. Stakeholder analysis is thus often the starting point for most participatory work.

CASE STUDY
KARU (Nigeria)
The city of Karu used a stakeholder analysis to identify and select the participants for a series of stakeholder consultation meetings organised as part of the CDS process in 2002. The approach to the analysis was developed by UN-HABITAT. The stakeholder analysis was conducted by a Consultation Organising Committee (COC) made up of representatives from community leaders, local government, and representatives of both the physical planning authorities and the private sector. Because the COC was comprised of people with indepth knowledge of the town and its inhabitants, the committee was perceived to be in a good position to identify stakeholders. One of the main goals of the CDS process in Karu was to develop viable approaches to ﬁnancing, operating, and maintaining public service delivery and infrastructure. The COC started by compiling a full list of those with a stake in, information about, and responsibility for providing urban services. These stakeholders were then assigned to three categories: private sector (informal and formal), public sector (local, state, and federal government), and the popular sector (civil society organisations, traditional leaders). The stakeholder analysis that resulted was conducted separately for each of the three categories, using a matrix of stakeholder responsibility (low-high) and stake (low-high). The analysis of public sector stakeholders focused on the statuary responsibility of each (for functions related to providing public services), whereas the analysis for the private sector and civil society focused more on the interests and resources needed in providing urban services. The stakeholder analysis was perceived as useful to the city because it provided a good overview of the stakeholders’ roles (by identifying the responsibilities of local versus state and federal, for example). The stakeholder analysis also revealed that private companies and individuals were investing in and providing key urban services (including water) through investments in boreholes and storage facilities.

108

LED RESOURCE GUIDE

Tips
Use local resources for ﬁrst cut at main groups. Although external consultants may be needed in some aspects of the local economy assessment, they are of little value in helping to identify and map local stakeholders. Local representative organisations can be useful because they are likely to have knowledge of and ties to the local economy. Cast a wide net. A broad range of stakeholders and institutions will likely play important roles in various aspects of the local economy and will have important inputs to the assessment and strategy development process. Therefore, it is important to look beyond the most obvious stakeholders. Make use of visual tools. Links between institutions are most easily identiﬁed through visual processes, such as drawing maps or diagrams that use symbols and physical distance to represent the nature and scope of relationships. Make the process iterative. As more information and new informants emerge, the picture of institutions and their relationships will change. This process must be allowed to evolve.

The ﬁrst step in stakeholder analysis is to compile a list of stakeholders. These can be individuals or wider groups and institutions, but for stakeholder groups, three broad types are relevant: the public sector (city or municipal departments, regional or national government departments, universities, and so forth), the private sector (such as chamber of commerce, chamber of handicrafts, banks, cooperatives, and research institutions), and nonproﬁt organisations (including minority associations, NGOs, trade unions, and women’s associations). Once the stakeholders have been identiﬁed, relationships between organisations and institutions are mapped out to understand roles, relations, and gaps in existing linkages among them. Stakeholder analysis is a versatile tool that can be used at different levels of formality. Although the analysis is often conducted as a facilitated group process, it may also involve administering a questionnaire to stakeholder organisations (business groups and neighbourhood organisations, for example) with questions concerning mission, geographical area of intervention, main types of activities, and other rele-

Further Information
For a step-by-step guide to the stakeholder analysis template developed by UN-HABITAT, see: http://hq.unhabitat.org/cdrom/ governance/html/st.htm. For a practical step-by-step guide for conducting a stakeholder analysis, see: http://www.scu.edu.au/schools/gcm/ar/arp/stake.html. For a checklist of possible stakeholders to include in a local economy analysis, see: http://www.icmm.com/library_pub_detail.php? rcd=183. For discussion of stakeholder analysis tools in the context of urban upgrading (based on GTZ and Norwegian Agency for Development Cooperation), see: http://web.mit.edu/urbanupgrading/upgrading/issues-tools/tools/Ident-stakeholders.html.
ANALYSING THE DATA

For a template used to prepare a stakeholder organisation questionnaire, download the International Labour Organisation’s LED Operational Guidelines in Post-Crisis Situations at: http://www.ilo.org/dyn/empent/empent.Portal?p_prog=&p_subprog=&p_category= TOOLS. For a list of tools used by the World Bank to conduct stakeholder and institutional analysis (designed for poverty and social impact analysis), see: http://lnweb18.worldbank.org/ESSD/sdvext.nsf/81ByDocName/ToolsandMethodsInstitutionalanalysis.

109 vant issues. This may be particularly relevant for larger economies where there are many stakeholder organisations. A common approach to pulling together the results of the stakeholder analysis is to draw a matrix of stakeholder inﬂuence (low-high) and interest (low-high stake). Based on this, it is possible to develop a strategy for involving each target person or group. Stakeholder analysis is a relatively easy tool to use, but it requires an experienced and trained facilitator who has good people skills and is sensitive to local cultural norms. This analysis does not require any resources other than data and human resources. But even though stakeholder analysis generally has a low resource intensity, it may be moderate to high if a survey requiring more formal analysis is used (such as a survey questionnaire that involves a large territory or population sample).

What Key Inputs Are Required for Stakeholder Analysis?
When a good overview of the local economy is available, and stakeholders are included in data collection, data for stakeholder analysis can be obtained. For this analysis, cities need: A list of main stakeholder groups. Names and contact details for stakeholders to be included in the analysis.

What Other Analytical Tools Are There?
The most widely used tools for local economy analysis have been described in this chapter. However, other tools can be useful as well. Some of these are briefly described in table 8.5.

T A B L E 8 . 5 Overview of Other Data Analysis Tools
Analytical tools
Locational Gini Index

Description
Uses the standard Gini coefﬁcient to measure the spatial dispersion of industry, by replacing individuals with regions. Can be calculated for territories and/or sectors. This index has become a standard measure but is not effective for crossregional comparisons.

Further Information
For an example of how the locational Gini is used to measure industry concentration in the European Union, see: http://www.hec.unil.ch/ mbrulhar/papers/tep955.pdf. For an example of how the locational Gini is used to measure industry concentration in China, see: http://rspas.anu.edu.au/economics/ publish/papers/wp2001/2001-07%20MeiWenWP2v.pdf. For an example of how entropy indexes are used in a cluster mapping exercise in Ireland, see: http://www.tcd.ie/iiis/documents/discussion/ pdfs/iiisdp89.pdf. For discussion on use of the Theil Index to measure inequality, see: The Young Person’s Guide to the Theil Index: Suggesting Intuitive Interpretations and Exploring Analytical Applications at: http://papers.ssrn.com/ sol3/papers.cfm?abstract_id=228703. For a set of guidelines on the use of gender analysis, see: http://www. worldbank.org/wbi/sourcebook/sba109.htm. For a range of approaches to gender analysis, see: http://www.ilo.org/ public/english/region/asro/mdtmanila/training/unit1/plngaps1.htm. For an example of how gender analysis is applied in the U.S. state of West Virginia, see: http://www.polsci.wvu.edu/ipa/par/report_ 13_2.html. For an introduction to various systems thinking methodologies, see the Arizona State University Business College paper at: http://www. public.asu.edu/~kirkwood/sysdyn/SDIntro/SDIntro.htm. For a comprehensive guide to systems thinking, see the Road Maps series developed by Massachusetts Institute of Technology’s Sloan School of Management at: http://sysdyn.clexchange.org/road-maps/ rm-toc.html. The Systems Dynamics Society Web site provides many resources and publications (for subscribers only): http://www.systemdynamics.org. For a mini-simulation exercise and a brief introduction to dynamics modelling at London Business School Systems Dynamics Group, see: http://www.london.edu/sysdn.html.

Entropy indexes
LED RESOURCE GUIDE

Measure the spatial dispersion of industry, as an alternative to the locational Gini index. Allow for weighting within subgroups and thus have the potential to decompose data by subgroups (for example, to explain the contribution of speciﬁc sectors to the overall geographic concentration).

Gender analysis

110

Involves a systematic analysis of the roles of women and men in the economy and the impacts of economic and social policy on women versus men. Can be a valuable tool for understanding human capital potential and the issues currently restricting economic participation, particularly among women. Can also be valuable for understanding the informal economy.

Systems thinking

An approach for analysing and managing complex feedback systems. Most appropriate for analysing interrelationships between different issues in the economy and anticipating development outcomes by mapping possible chains of causes and effects; fairly complex technique that requires both good understanding of economic issues and facilitation skills, and is ideally aided by developing a computer simulation model in which causes and effects can be tested.

Similar Documents

Popular Essays