Relative Political Capacity: New Models & Data Updates Marina Arbetman Rabinowitz Kristin Johnson Claremont Conference on Political Economy Indicators Indicators of Political Development Indicators of Well being: Electricity, calories per capita, education, health, etc. Deutsch (1966), Rokkan (1970), Gurr (1974), World Handbook of Social Indicators (WB), Human Development Index (UN), Millennium Development Goals. Democracy Research: Participation, representation, electoral choice, institutions, bureaucracy (Campbell, Converse, Stokes and Miller, 1960; Verba, Nie and Petrocik, 1972; and Fiorina, 1981). Freedom House Index, Polity IV (democracy/ autocracy), Governance Indicators. Claremont Conference on Political Economy Indicators Theoretical issues These approaches each fall into the trap of measuring the consequences or reflecting phenomena other than capabilities but have the advantage of measuring the phenomena directly. Measures of well being reflect government and elite choice in resource allocation Measures of governance favor democratic systems of government over other forms of governance. Claremont Conference on Political Economy Indicators Relative Political Capacity Measures the discretionary power of the government to implement a desired policy. The Relative Political Capacity of a government is the ratio of revenues a government extracts, compared to predicted levels a society could attain based on economic endowment. RPC = Actual Government Revenue Predicted Government Revenue Claremont Conference on Political Economy Indicators Relative Political Capacity RATIO OF REVENUE TO GDP Boundary of High Performance Predicted Performance Successful States Fragile States Boundary of Low Performance Poor Developing Developed Claremont Conference on Political Economy Indicators Examples: RPC across time Bolivia High RPC Thailand Botswana Angola 1960 1970 1980 2000 2005 1990 1970 1960 1970 1980 1980 2005 2000 1990 1960 1980 1990 2000 2005 1970 1990 Low RPC 2005 2000 0 2000 4000 GDP per Capita 6000 8000 Claremont Conference on Political Economy Indicators RPC 2000 positive Lesotho Brazil Hungary Barbados Poland Israel Portugal Sweden Austria Burkina Faso Tunisia Ecuador 0 negative New Zealand Denmark Uruguay Kenya Finland Morocco Malta Greece Singapore Papua New Guinea USA Indonesia South Korea Jordan Fiji Malaysia Sri Lanka Thailand Bolivia Philippines Mauritius Iran Argentina Kuwait Bangladesh Gabon Paraguay Haiti Myanmar RPC 0 Income Claremont Conference on Political Economy Indicators 40,000 RPE General Model Yit = α + β Xit + Vr Yit = Adjusted tax revenue for country i at time t X it = Vector of variables that determine potential tax collection Vr= White noise disturbance Claremont Conference on Political Economy Indicators Inter country comparisons: Models for Developing Societies Model 1: Tax/Gdp= α + β1 time - β2 min/gdp + β3 gdpCap + β4 exports/gdp + β5 Oil + ξ Tax= (TaxRev-NonTax-SocSec)/Gdp RPC1 = Real /Predicted Note: Control for expenditures Model 2: Tax/Gdp= α + β1 time - β2 min/Gdp + β3 agr/Gdp + β4 exports/Gdp + β5 Oil + ξ Tax= (TaxRev-NonTax-SocSec)/Gdp RPC2 = Real /Predicted Claremont Conference on Political Economy Indicators Inter country comparisons: Models for Developed Societies Model 3: Tax/Gdp= α + β1 time - β2 min/gdp + β3 gdpCap + β4 Exports/gdp + β5 health/gdp + ξ Tax= (TaxRev-NonTax-SocSec)/Gdp RPC3 = Real /Predicted Claremont Conference on Political Economy Indicators Empirical Inventory National Level model Unit of analysis: country Number of countries in the sample: 129 Time span: 1960-2005 Notes: Sample includes Poland 86-on, Hungary 89-on, Romania 88 on. These countries use NMP in their National Accounts before those dates so the data is not comparable to GNP/GDP methodology). The sample includes China. Claremont Conference on Political Economy Indicators Variables & Sources country Country Name year 1960 - 2004 Total revenue divided by GDP*: Main Source GFS, IMF 1972-2005; World Bank 1960 – 1972, Where data was missing we used overlapping series and National Sources. totrev nontax Non tax revenue divided by GDP*. Main Source GFS, IMF 1972-2005; World Bank 1960 – 1972, Where data was missing we used overlapping series and National Sources. Social security revenues divided by GDP*. Main Source GFS, IMF 1972-2005; World Bank 1960 – 1972, Where data was missing we used overlapping series and National Sources. Agriculture GDP divided by total GDP*. Main Source GFS, IMF 1972-2005; World Bank 1960 – 1972, Where data was missing we used overlapping series and National Sources. Exports divided by total GDP*. Main Source GFS, IMF 1972-2005; World Bank 1960 – 1972, Where data was missing we used overlapping series and National Sources. Mining production GDP divided by total GDP*. Main Source GFS, IMF 1972-2005; World Bank 1960 – 1972, Where data was missing we used overlapping series and National Sources. Real GDP per capita in constant dollars. Main Source: Penn World Tables, Summers & Heston, additional sources include the IMF GFS & World Bank, World Development Indicators. Where data was lacking we used national sources. Totrev – nontax - socsec socsec agri exp mining Realgdp taxratio Oil Oil Exports/total exports. Main Source: Direction of Trade Statistics. Claremont Conference on Political Economy Indicators Data Collection Issues Data relies on national government accounts Variances in tax reporting and accounting are noted by IMF or WB, in most cases. Currency changes and differences in reporting timing in high inflation situations need to be normalized. National sources may be necessary to supplement Claremont Conference on Political Economy Indicators Measurement & Conceptualization Health expenditures: before 1990s mostly government reported public expenditures, now private are part of the mix in both developed and developing but not always reported Social Security: 1) some countries have privatized part of SS. 2) The argument to exclude social security has been that those transfers have been allocated previous to the collection and therefore there is little room for any further manipulation by the government. This is not the case, those funds are not untouchable, and the deficit can be manipulated by Central Banks or Congress Agriculture: Still our best empirical fit for developing countries (not for developed) but 1960s is the anchor so we need to control for time. Taxation: Indirect vs. direct taxation Claremont Conference on Political Economy Indicators Sub National Models Yit = α o o + Yit = Adjusted tax revenue for country i at time t X it = Vector of variables that determine potential tax collection Vr= White noise disturbance β Xit + Vr Conceptually, the same model as national level. Constraints to choosing the variables are country specific Claremont Conference on Political Economy Indicators Relative Provincial Political Capacity (RPPC) Model 1: Transf/gdp = α + β1 time + β2 min/gdp + β3 agr/gdp + β4 rev/gdp + ξ Model 2: Transf/gdp= α + β1 time + β2 gdpCap + β3 rev/gdp + β4 min/gdp + ξ Model 3: Rev/gdp= α + β1 time + β2 transf/gdp + β3 agr/gdp + β4 min/gdp + + β5 Subsidies+ ξ Claremont Conference on Political Economy Indicators Empirical Inventory Sub National Level Model Unit of analysis: province/state/region Time span: Ideally 1990 – 2005, dependant on individual country Bolivia, Brazil, China, EU, India, Indonesia, Iran, Mexico, Nigeria, Thailand, Sudan, USA, Claremont Conference on Political Economy Indicators Data Collection In many countries the transfers are decided according to a Congressional Allocation Funds formula. If the allocations are decided on annual negotiations (1), the variable captures the concept of RPPC. If the formula is inflexible (2), the flexibility comes from the allocation of central government expenditures to the provinces. It is also important to know that the General Appropriation Act does not show if or how much of the money is released, so instead of following to entry number (Transfers, Subsidies. Models 1& 3) in some cases is better to look at the exit number (Expenditures. Model 2) Claremont Conference on Political Economy Indicators Variables: RPPC province state Year Code Split province name State name 1980-2005 2 digit codes except for split provinces (3rd digit added for identification) 1 or 0 for provinces that have not split Population Totrev number of people reported in census * Includes 1.Surplus + 2. Local Gov Original Receipts + 3. Balanced Budget + 4. Local Gov. Loans Localrev Transfers Taxes collected by the local government from the local population transfers, from General Allocation Fund Transf_gdp Expend_gdp Grdp constant YEAR Grdp market agriculture mining Oil Production RPC1 RPC2 Transfers divided by total GDP* Expenditures divided by total GDP* gross regional domestic product in constant YEAR prices in local currency gross regional domestic product at market prices in local currency Agriculture GDP divided by total GDP* Mining production GDP divided by total GDP* Dummy or % oil produced/Total country production controlling for agriculture (specify all variables used in the model) controlling for gdp per capita (specify all variables used in the model) Claremont Conference on Political Economy Indicators Sources Examples: RPPC Expend_gdp Main Source: Statistical Center of Iran, available at: http://www.sci.org.ir/english/default.htm agriculture Main Source: Statistical Center of Iran, available at: http://www.sci.org.ir/english/default.htm Main Source: Statistical Center of Iran, available at: http://www.sci.org.ir/english/default.htm Main Source: Economic & Financial Review, The Central Bank of Nigeria 1979 – 1987; ISSN: 008-9281; Central Bank of Nigeria, Annual Report & Statement of Accounts. 1980 – 2005. ISSN: 0069-1577 Main Source: Economic & Financial Review, The Central Bank of Nigeria 1979 – 1987; ISSN: 008-9281; Central Bank of Nigeria, Annual Report & Statement of Accounts. 1980 – 2005. ISSN: 0069-1577. When data was lacking, overlapping series were obtained from FAO Statistics available from the UN. mining Rev_gdp agriculture mining Main Source: Economic & Financial Review, The Central Bank of Nigeria 1979 – 1987; ISSN: 008-9281; Central Bank of Nigeria, Annual Report & Statement of Accounts. 1980 – 2005. ISSN: 0069-1577 Claremont Conference on Political Economy Indicators Applications International Conflict Organski & Kugler 1980; Lemke 1996; Tammen et al 2002. Kugler et al 1997; Benson & Kugler 1998; Johnson 2007. Arbetman, Kugler & Organski 1980, 1994, 1999; Feng Kugler & Zak 2000, 2007. Exchange Rates: Arbetman 1990; Growth: Leblang 1997; Inflation: Alcazar 1997; Private Investment: Feng & Chen 1997, Feng 2004; Policy Implementation: Snider 1997; Informal Markets: Arbetman 1990, 1994; Arbetman & Ghosh 1997. Rouyer 1997; Swaminanthan 2005; 2007; Johnson 2007. Claremont Conference on Political Economy Indicators Internal Conflict Demographic Transitions Economic Applications Sub National Relative Political Capacity RPC at different levels of development + Very High High Average Predicted Performance Low Very Low _ Poor Developing Developed Claremont Conference on Political Economy Indicators
Comments
Report "Arbetman-Rabinowitz Johnson 2007 Relative Political Capacity"