Free Essay

Propensity Score Matching

In:

Submitted By dante042
Words 1537
Pages 7
Breves comentarios respecto al desarrollo del Propensity Score Matching

Introducción.
Dentro de los métodos de evaluación de impacto a partir de la familia de técnicas de “matching”, el Propensity Score Matching (PSM) es uno de los más utilizados por los investigadores, entre otras cosa por los supuestos para abordar los llamados sesgos de selección para la estimación de efectos causales por intervención.
Al respecto existen algunos puntos de discusión concernientes a su desarrollo e implementación, los cuales son pocamente discutidos o abordados en trabajos empíricos; por lo que se plantean a continuación algunos de los aspectos más relevantes que deben ser considerados conceptualmente, al momento de llevar a cabo una evaluación utilizando éste método.
Diferencias entre los modelos logit y probit
El Propensity Score Matching (PSM) es estimado utilizando un modelo de regresión logit o probit, que son modelos de probabilidad no lineal para calcular la probabilidad que tiene un individuo de recibir el tratamiento de acuerdo a una serie de características socioeconómicas, geográficas, entre otras. Sin embargo, la aplicación de cualquiera de ellos responde a supuestos subyacentes y a las propiedades de la información con la que se cuenta.
Los modelos probit mantienen como supuesto básico la distribución normal de las observaciones, mientras la distribución como una curva logística determina el uso de los modelos logit. Los resultados estimados no difieren mucho entre sí, pero son las diferencias operativas las que han potenciado la aplicación del modelo logit en la mayoría de los estudios empíricos, en virtud de su menor complejidad de manejo. Mientras los modelos probit presentan una mayor complejidad en el cálculo de la función de distribución normal ya que sólo puede calcularse en forma de integral.
Pruebas de balanceo
Una vez estimado el Propensity Score, es preciso verificar el balance de las covarianzas para los grupos de tratamiento y control. Rosembaum y Rubin (1983), indican que el balanceo es de suma importancia para eliminar los sesgos, debido a que el PSM tiene la capacidad de eliminar los diferentes sesgos asociados a las diferencias existentes antes de aplicar el tratamiento entre el grupo de tratamiento y de control.
Un tema fundamental surge cuando, una vez estimado el propensity score, es conveniente verificar los supuestos del modelo de regresión y si éstos empatan o no con las pruebas de balanceo. Dicho tema no se explicita en la literatura sobre el tema, no obstante se encuentran referencias recurrentes sobre los criterios seguidos en este punto.
De acuerdo a D´Agostino y Rubin (2000), el propensity score, per se, sirve sólo como un vehículo para balancear la distribución de covariables observada a través de los grupos de tratamiento y control. Por tanto, el éxito de las estimaciones del propensity score, consiste en evaluar el balance resultante en vez del ajuste del modelo usado para crear las estimaciones del propensity score.
En diversas evaluaciones de impacto se muestra el proceso de implementación del PSM, y de acuerdo a los resultados expuestos, aún cuando existen variables no estadísticamente significativas, se mantienen en el propensity con el objetivo de evitar sesgos en el resto de los coeficientes a causa de variables omitidas, ya que el conjunto de todos los coeficientes pueden ser altamente significativos. Salvo esta enunciación, no se brindan explicaciones más esclarecedoras o en su defecto, no se brindan elementos concluyentes y se da por sentado esta forma de proceder.
Respecto a este punto, Heckman et al (1997) sugieren que la omisión de variables importantes puede incrementar el sesgo en la estimación, pero se deben incluir sólo las variables que influyen en la probabilidad de ser tratado y en la variable de resultado, y que a su vez se ven afectados por el programa. Y al decir de Rubin y Thomas (1984), a menos que una variable pueda ser excluida porque existe consenso de su irrelevancia para el estudio o es una covariable inapropiada, es aconsejable incluirla en el PSM, incluso si no es estadísticamente significativa.
Adicionalmente, Lechner y Smith (2002) comentan que la inclusión de variables no significativas no sesga las estimaciones o las hace inconsistentes, sino que su efecto es que se incrementa la varianza en las estimaciones.
Por otro lado, Byryson et al., (2002) recomiendan no sobreparametrizar los modelos, ya que incluir demasiadas variables en el modelo usado para caracterizar la probabilidad de recibir tratamiento, reducirá la probabilidad de hallar una zona de soporte común.
Los elementos anteriores llevan a la consideración de que es pertinente la implementación de las pruebas de balanceo y asegurar los correctos resultados derivados de ellas, aún cuando las pruebas de los modelos probabilísticos se no se cumplan a cabalidad.
Inconsistencias en la implementación del Propensity Score Matching
De acuerdo con Austin (2007) en un estudio sobre la implementación de los métodos de propensity en 47 estudios en la literatura médica entre 1996 y 2003, existen dos procedimientos una vez estimado el PSM mediante un logit o probit. Evaluar el balance entre las covariables de tratamiento y control y estimar el efecto de tratamiento en las variables de impacto en el PSM.
Los métodos propuestos para evaluar el balanceo incluyen la estandarización de diferencias, comparar momentos de orden superior, estadísticos del propensity score y gráficas de cuantiles para cada variable.
Jennifer Hill (2008) retomó dicho estudio e identificó ciertos problemas recurrentes en la literatura de PSM en la cual existe una mala interpretación para su aplicación y por lo tanto se implementa sin tomar las debidas precauciones. Dichas inconsistencias se refieren a: * Una incorrecta definición de los contrafactuales, como una variable que predice el tratamiento y el resultado. Su mala interpretación ha conducido a describirla como variables que predicen la asignación del tratamiento y los predictores en la estimación del modelo de propensity score. Esta práctica ha originado el uso de la regresión paso a paso y verificar el balanceo sólo en las variables restantes en la estimación del modelo final * Además del balanceo, deben implementarse otras pruebas, tales como: momentos de orden superior, la varianza y covarianza. Dado que puede presentarse desbalanceo en órdenes superiores, aún cuando en la media se haya logrado balancear la muestra. * Se debe optar por un análisis de sensibilidad de las estimaciones debido a que no existe una teoría concluyente sobre el balanceo que indique hasta qué punto es suficiente concluir. * Verificar si se está realizando inferencia sobre la muestra o sobre la población. * Errar en el diseño de la dirección de la dependencia en el PSM conduce a que los errores estándar sean muy grandes.
Diferentes tipos de pruebas de balance
De acuerdo a Wang-Sheng (2006) existen discrepancias en la literatura en torno a las pruebas de balanceo, ya que coexisten diferentes tipos que derivan en diferentes resultados. Por tanto, es importante distinguir entre las diferentes pruebas de balanceo conducidas antes y después del matching.
Las pruebas de balanceo son un medio para un fin, y pueden ser consideradas útiles sólo si las pruebas de balanceo permiten estimaciones del efecto de tratamiento insesgadas. (Wang, 2006).
La confusión en la literatura se ha presentado porque el término “pruebas de balance” fue implementado como la prueba de Dehejiaand Wahba (DW test) y para verificar el balanceo en el match de la muestra. En la literatura, las pruebas de balanceo conducidas antes del match fueron introducidas por Rosenbaum and Rubin (1984), y aplicadas en Rubin (1997), Dehejia y Wahba (1999, 2002), Michalopoulos, Bloom y Hill (2004), y Dehejia (2005). Mientras las pruebas de balance conducidas después del match fueron nombradas por Smith y Todd (2005) como pruebas de balance.
La prueba de balanceo antes del match constituye más una prueba de especificación, o bien una prueba de balanceo, pero cuando se realiza la estratificación del propensity score. Debe ser distinguida de una prueba de balanceo después del match, que se utiliza realmente para verificar si el match con el grupo de control puede ser considerado representativo del contrafactual. Las diferencias en resultados se deben a las diferencias en el uso de las pruebas de balanceo antes o después del match
Las pruebas de balanceo pueden categorizarse en de acuerdo al test de las propiedades sobre toda la muestra o después de haber aplicado el algoritmo de “matcheo” y a unidades de la muestra

Bibliografía
Bryson, A., Dorsett, R. y Purdon, S. (2002). The use of propensity score matching in the evaluation of active labour market policies. Policy Studies Institute and National Centre for Social Research.
Heckman, J., Ichimura, H. y Todd, P. (1997). Matching As An Econometric Evaluation Estimator. Review of Economic Studies (1998) 65, 261-294.
Huber, Martin, Testing for covariate balance using quantile regression and resampling methods, University of St. Gallen, Dept. of Economics.
Medina Moral, Eva, Modelos de elección discreta, 2003.
Rosenbaum, P. y Rubin, D. (1983) The Central Role of the Propensity Score in Observational Studies for Causal Effects. Biometrika, Vol. 70, No. 1. (Apr, 1983), pp. 41-55.
Wang-Sheng, Lee. (2006). Propensity Score Matching and Variations on the Balancing Test, Melbourne Institute of Applied Economic and Social Research, The University of Melbourne.

--------------------------------------------
[ 1 ]. Discrepan, únicamente, en la rapidez con que las curvas se aproximan a los valores extremos, la función logística es más achatada que la normal al alcanzar, esta última, más rápidamente los valores extremos (0 y 1).

Similar Documents

Premium Essay

Juvenile Offenders

...Juvenile offenders November 25, 2013 Juvenile offenders Juvenile offenders are categorized as offenders aged below the age of 18 years. The option of transferring these offenders to adult jurisdictions has been ongoing since the inception of these courts. Various reasons has contributed to this including inadequacy of punishment on serious offenses and the propensity to continue offending after release also called recidivism among others. Various researches have been carried out on this aspect of transferring juvenile to adult courts c with different recommendations. This discussion, reviews an article, “Differential Effects of Adult Court Transfer on Juvenile Offender Recidivism” by Loughran et al. (2010) in the law and human behavior journal. The article sought to investigate on the effects on the ability of the juveniles to repeat offenses if they are taken to adult courts. It is expected that juveniles below the age of 14 years are tried by juvenile courts and not adult jurisdiction. The article notes that there are null effects of transfer on re-arrest (Loughran et al. 2010). Most of the juveniles that may be taken to adult courts are not prone to recidivism therefore it means that even if the juveniles are taken to the adult courts, there is less differential in their behaviors on their release. Even though, various studies have indicated that juvenile offenders that are transferred to adult courts have higher...

Words: 661 - Pages: 3

Premium Essay

Power Point 4ps

...School of Economics Diliman, Quezon City UPSE Discussion Papers are preliminary versions circulated privately to elicit critical comments. They are protected by Republic Act No. 8293 and are not for quotation or reprinting without prior approval. The Impact of Philippines’ Conditional Cash Transfer Program on Consumption1 Melba V. Tutor* *University of the Philippines School of Economics; mvtutor@up.edu.ph Abstract Pantawid Pamilyang Pilipino Program provides cash grants to poor households conditional on pre-determined investments in human capital. This study analyzed its impact on consumption using the 2011 Annual Poverty Indicators Survey. Average treatment effect on the treated (ATT) is estimated through propensity score matching methodology. Heterogeneous impacts are examined among the bottom 20% of income distribution. The study finds that among the total sample, per capita total expenditures is not affected by the program. In per capita monthly terms, only carbohydrates and clothing significantly increased. As expenditure shares, education and clothing registered significant positive impact. No impact is observed on health spending, both in per capita terms and as a share of expenditure. The impact of Pantawid Pamilya on consumption is more pronounced among the poorest fifth of households. Results show that households have responded to program conditionalities but there is very little room to improve consumption of other basic needs. The...

Words: 9006 - Pages: 37

Free Essay

Crowdfunding

...IMPLEMENTATION OF PROPENSITY SCORE MATCHING Marco Caliendo IZA, Bonn Sabine Kopeinig University of Cologne Abstract. Propensity score matching (PSM) has become a popular approach to estimate causal treatment effects. It is widely applied when evaluating labour market policies, but empirical examples can be found in very diverse fields of study. Once the researcher has decided to use PSM, he is confronted with a lot of questions regarding its implementation. To begin with, a first decision has to be made concerning the estimation of the propensity score. Following that one has to decide which matching algorithm to choose and determine the region of common support. Subsequently, the matching quality has to be assessed and treatment effects and their standard errors have to be estimated. Furthermore, questions like ‘what to do if there is choice-based sampling?’ or ‘when to measure effects?’ can be important in empirical studies. Finally, one might also want to test the sensitivity of estimated treatment effects with respect to unobserved heterogeneity or failure of the common support condition. Each implementation step involves a lot of decisions and different approaches can be thought of. The aim of this paper is to discuss these implementation issues and give some guidance to researchers who want to use PSM for evaluation purposes. Keywords. Propensity score matching; Treatment effects; Evaluation; Sensitivity analysis; Implementation 1. Introduction Matching has become a popular...

Words: 20722 - Pages: 83

Free Essay

Termpaper

...Assignment 1 Yang Liu May 5, 2015 Assignment 1, Part 1 (1) Table 0.1: Estimate a logit using solver Product of probability Log likelihood Intercept Eduation coefficient Age coefficient 2.06641E-11 -24.60262143 -11.15550863 0.531907452 0.113507304 (2) M EE d = M E Ag e = βE d e X β (1 + e X β )2 β Ag e e X β (1 + e X β )2 = 1 N βE d (i ) e X βi Σ N 1 (1 + e X βi )2 = 1 N β Ag e(i ) e X βi Σ N 1 (1 + e X βi )2 1 Result: Table 0.2: Add caption Marginal Edu effect Marginal age effect 0.085732302 0.018294992 (3) Table 0.3: linear probability model Coefficients t Stat -1.541134276 0.097414096 0.020646694 Intercept EDUC AGE Std. Err. 0.429125046 0.025225082 0.005153578 -3.591340779 3.861794961 4.00628341 (4) H = −Σi p i (1 − p i )x i x i = X ΩX The diagonal of matrix Ω equal to p i (1 − p i ) The Hessian matrix   8.058949138 105.838015 300.4357025 H =  105.838015 1434.435526 3840.728111 300.4357025 3840.728111 12190.99391 (0.1) The variance-covariance matrix equals to the inverse of Hessian matrix Table 0.4: variance-covariance matrix Constant Constant Edu Age Edu Age 10.2699517 -0.511915922 -0.091816163 -0.511915922 0.029972725 0.003172894 -0.09182 0.003173 0.001345 2 (5) According to Student’s t distribution, under 49 freedom degree, the probabilities that these coefficients are 0s are less than 0.005. We can say these...

Words: 856 - Pages: 4

Premium Essay

Healthcare

...International Financial Reporting Standards (IFRS) in Europe. First, analyzing the determinants of voluntary IFRS adoption by publicly traded German firms during the period 19982004, we find that size, international exposure, dispersion of ownership, and recent IPOs are important drivers. Second, using the results from this determinant model to construct propensity score-matched samples of IFRS and German-GAAP (HGB) firms, we document significant differences in terms of earnings quality: IFRS firms have more persistent, less predictable and more conditionally conservative earnings. Third, analyzing information asymmetry differences between IFRS and HGB firms, we show that IFRS adopters experience a decline in bid-ask spread of 70 base points and an average of 17 more days with price changes per year. On the other hand, IFRS adopter’s stock prices seem to be more volatile. In the light of some important limitations of our study, we discuss IFRS-related research opportunities in post-2005 Europe. Keywords: IFRS, earnings quality, earnings attributes, information asymmetry, standard setting, IAS Regulation, Europe, propensity-score matching,...

Words: 11298 - Pages: 46

Premium Essay

Medical Out-of-Pocket Expenses

...Medical Out-of-Pocket Expenses, Poverty, and the Uninsured∗ Kyle J. Caswell† and Brett O’Hara SEHSD Working Paper 2010-17‡ U.S. Census Bureau Washington, D.C. December 29, 2010 Abstract The National Academy of Sciences (NAS) Panel on Poverty and Family Assistance argued that the current official U.S. poverty measure should be updated to capture changes in the population’s healthcare costs and needs; families with sufficiently high medical out-ofpocket (MOOP) expenditures may be ‘poor’ even though they are not counted as such. This research offers three distinct advances toward achieving the NAS recommendations as they concern MOOP spending. Firstly, this paper uses the newly collected MOOP expenditure data from the CPS ASEC, and analyzes its quality vis-` -vis alternative sources. Secondly, a poverty estimates that incorporate the MOOP spending data from the CPS ASEC are produced in such a way as to be consistent with the NAS recommendations. These direct estimates are an improvement over previous estimates, conditional on obtaining high-quality data, because modeling MOOP expenditures from other surveys is not needed. Third, this paper investigates how the distribution of MOOP expenditures, and the poverty estimates, change when it is assumed that the uninsured have the spending patterns of the insured. The main results are: 1) the new MOOP expenditure data is high quality; 2) incorporating observed MOOP expenditures increases the incidence of poverty across the population...

Words: 4481 - Pages: 18

Free Essay

Student

...many articles is too complex for display using the software we use to publish SP, so we are experimenting with PDFs. The articles span a broad spectrum, including the evaluation of bias in a nonprobability sample, the review of assumptions in a nonprobability sampling method that provide the potential for bias, the conditions under which a nonprobability sampling design can lead to valid conclusions in comparative research, case studies on the use of nonprobability methods and samples to facilitate a probability-based study, and a proposed method to combine probability and nonprobability samples under certain conditions. Gerty Lensvelt-Mulders and colleagues use a probability-based web survey with telephone follow-up and propensity score matching in order to evaluate bias in a nonprobability web panel survey. This design and analytic approach allow them to attempt to separate bias due to self-selection from bias due to undercoverage in the panel survey. Although not nearly as much in the survey literature, Respondent Driven Sampling has received considerable attention as a nonprobability sampling method that claims to produce representative estimates. In her article, Sunghee Lee dissects the sources of error in RDS from a total survey error perspective. Murray Straus presents an evaluation of the validity of cross-national comparisons using nonprobability sampling, when the study design is held similar across countries. Based on data from 29 countries, he uses...

Words: 591 - Pages: 3

Premium Essay

The Impact of Internet Access at Home and/or School on Students’ Academic Performance

...Anísio Teixeira” (National Institute of Educational Research Anísio Teixeira) or INEP for years 2007 and 2009. Using the propensity scores matching methodology the paper assesses the impact of internet on standardized Portuguese and mathematics tests that students in all public schools take in the 8th grade while controlling for an extensive list of variables which includes various household assets, parental education and involvement with student activities, the school physical infrastructure, safety, electronic equipments and library as well as teachers’ academic history, experience and pedagogical strategies. Following Lechner (2001) the multiple treatment case is examined constituting of four alternatives: students with no Internet access at home or at school, students with Internet access at home but not at school, students with Internet access at school but not at home, and students with Internet access at home and at school. The results suggest that Internet access at home considerably improves test scores of 8th graders. The results hold for both Portuguese and mathematics tests and for both years 2007 and 2009. On the other hand, the Internet access at school either does not improve the academic performance of students (as it is the case for Portuguese test scores) or the effect is of much smaller magnitude (as it is the case for mathematics test scores) in 2007. In 2009 the impact of Internet access only at school is positive and significant though still of smaller magnitude...

Words: 437 - Pages: 2

Free Essay

Export Behavior and Firm Productivity in German Manufacturing

...Export Behavior and Firm Productivity in German Manufacturing A firm-level analysis Jens Matthias Arnold* and Katrin Hussinger** Abstract This paper examines the causal relationship between productivity and exporting in German manufacturing. We find a causal link from high productivity to presence in foreign markets, as postulated by a recent literature on international trade with heterogeneous firms. We apply a matching technique in order to analyze whether the presence in international markets enables firms to achieve further productivity improvements, without finding significant evidence for this. We conclude that high-productivity firms self-select themselves into export markets, while exporting itself does not play a significant role for the productivity of German firms. Keywords: Exports; Export-led growth; Total Factor Productivity; Heterogeneous firms. JEL-Classification: F10, F13, F14, D21, L60 Addresses: *Bocconi University, Milan, Italy (Email: jens.arnold@uni-bocconi.it) ** Centre for European Economic Research (ZEW), P.O.Box 10 34 43 68304 Mannheim, Germany. Email: hussinger@zew.de (corresponding author) 1 Introduction Why do some firms in an industry export, while others in the same industry persistently serve the domestic market only? What are the determinants behind these different patterns within sectors? How are these differences in export behavior related to productivity differences among firms? Do the best...

Words: 8781 - Pages: 36

Premium Essay

Exploring My View on Vaccination Issues

...With reference to the thousands of family’s affected by reactions to vaccines and are not entitled to compensation as families in other countries are: 1) I think there should be avenues for support if a child is permanently disabled as a direct result of a vaccination (by participating in the vaccination program they are assisting in the maintenance of the herd immunisation of our country after all) 2) The numbers of any reaction reported and associated with medicine (including vaccines) in Australia are recorded at http://apps.tga.gov.au/PROD/DAEN/daen-report.aspx . These reports do not confirm causality but are a point for adverse reactions to be recorded from which the government can follow up on the safety of the reported medicines. Anyone can submit reports to this data base and the following information is from the TGA website: “Each year the TGA receives more than 17,000 reports of suspected adverse events to medicines and vaccines. In 2013 about 55% of these reports came via pharmaceutical companies, 17% from state and territory health departments, 10% from hospitals and hospital pharmacists and the rest from community pharmacists, general practitioners and consumers. The TGA asks for contact details from people making reports so that it can seek further information about suspected adverse events. The TGA does not accept anonymous reports. If you have any concerns about an adverse event it is important to also speak to a health professional.” With reference to comment...

Words: 631 - Pages: 3

Free Essay

Most Harmless Econometrics

... . . . . . . . . . . . . . . . . . Asymptotic OLS Inference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Saturated Models, Main E¤ects, and Other Regression Talk . . . . . . . . . . . . . . . Regression and Causality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 3.2.2 3.2.3 The Conditional Independence Assumption . . . . . . . . . . . . . . . . . . . . . . . . The Omitted Variables Bias Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bad Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Heterogeneity and Nonlinearity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Regression Meets Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii iv 3.3.2 3.3.3 3.4 CONTENTS...

Words: 114745 - Pages: 459

Free Essay

Ckd Education

...The Effects of Self-Management on the Progression of Chronic Kidney Disease Keywords: Chronic kidney disease, self-management, patient education, end-stage renal disease The Effects of Self-Management on the Progression of Chronic Kidney Disease Kidney disease is one of the leading causes of death in the United States and was the 8th leading cause in 2010 (Arialdi M. Miniño, 2012). Approximately 20 million U.S. adults who are over 20 years of age have chronic kidney disease (CKD), and most of them are unaware of their condition ("CDC, National chronic kidney disease "). This number continues to grow every year due to an increase in the incidence of hypertension, diabetes, obesity, and the aging population. CKD can develop into end-stage renal disease (ESRD) rapidly without intervention, and the only treatment currently available is renal replacement therapy or kidney transplant. The cost of treating those in the United States with ESRD in 2009 was approximately $40 billion dollars, and this cost continues to grow each year ("National Kidney and Urologic Diseases Information Clearinghouse (NKUDIC) "). Moreover, patients who develop kidney disease pay an even higher price with their life. Studies have shown that CKD is associated with a higher rate of cardiovascular disease as well as all other related mortalities (Kurth, de Jong, Cook, Buring, & Ridker, 2009). An adult with CKD is 16-40 times more likely to die before reaching ESRD than an...

Words: 2015 - Pages: 9

Premium Essay

Death

...Enhancing Access and Control to Sustainable Livelihood Assets of the Manobo Tribe through Improved and Strengthened Selfgovernance of the Ancestral Territory – Effectiveness Review Full Technical Report Photo credit: PBPF Oxfam GB Women’s Empowerment Outcome Indicator June, 2012 Table of Contents Executive Summary...................................................................................................................... 1 1.0 Introduction and Purpose ....................................................................................................... 2 2.0 Intervention Logic of the Enhancing Access and Control to Sustainable Livelihood Assets Project ......................................................................................................................................... 3 3.0 Impact Assessment Design...................................................................................................... 5 3.1 Limitations in Pursuing the ‘Gold Standard’ ................................................................................. 5 3.2 Alternative Evaluation Design Pursued......................................................................................... 5 3.3 Intervention and Comparison Villages Surveyed .......................................................................... 7 4.0 Methods of Data Collection and Analysis ................................................................................ 8 4.1 Data Collection...

Words: 13369 - Pages: 54

Premium Essay

Low Voter Turnout Pros And Cons

...education cause political participation. Rather than conferring with the scholars, Kam and Palmer (2008:612) take on a differing perspective and focus on other factors that could in turn influence political participation. They find that “the same factors that propel individuals to pursue higher education also appear to propel them to participate in politics” (Kam, Palmer 2008:613). This ties into the arguments that there are underlying factors at work in the relationship. However, Kam and Palmer (2008:612) take quite a stand in opposing the casual relationship by claiming that “higher education is a proxy for pre-adult experiences and influences, not a cause of political participation.” This is because they found that “after the propensity-score matching process takes into account pre-adult experiences and influences in place during the senior year of high school, the effects of higher education per se on participation disappear” (Kam, Palmer 2008:612). While they only study the jump between high school and higher education, they do claim that education does not cause participation due to their results. Therefore, the scholars negate any causal relationship between educational levels and political participation. Likewise, a study in Norway on the length of compulsory education and voter turnout supports the view that political activity of people is driven by factors other than education. Furthermore, it demonstrates that increases in the level of education through school reforms...

Words: 578 - Pages: 3

Free Essay

Does It Pay Firms to Register for Taxes? the Impact of Formality on Firm Profitability

...Idee Die Analyse bezieht sich auf Daten einer Umfrage zur Produktivität bolivianischer Unternehmen die unter Mikro- und Kleinunternehmen im März 2007 durchgeführt worden ist. Dabei beschränkt sich die Erhebung auf Unternehmen innerhalb eines 10 km Radius zum Stadtzentrum und zum Finanzamt der vier größten Städte Boliviens: La Paz, El Alto, Santa Cruz und Cochabamba. Die daraus resultierende Stichprobe setzt sich aus 469 Unternehmen zusammen. Diese Erhebung beschränkt sich hierbei auf sechs Branchen, die einen großen Teil der selbstständigen Unternehmer und die allgemeine Wirtschaftsstruktur in Bolivien widerspiegeln. Dementsprechend wurde zwischen Unternehmen der Lebensmittel- und Bewirtungsindustrie, des Personen- und Frachttransports, der Bekleidungsproduktion aus Gewebe und Wolle, der Bekleidungsproduktion aus Wolle der Schwielensohler und zwischen Unternehmen der Holzfertigungen von Möbel unterschieden. Die Stichprobe basiert ferner auf Daten des geographischen Informationssystems, die wiederum auf einer Vollerhebung aller wirtschaftlichen Betriebe in den benannten Städten im August 2005 beruhen. Durch die Daten der Transportindustrie zur Registrierung von Firmenwagen wurden jene zu einer umfassenden Auswahlgrundlage für städtische Gebiete ergänzt. Die Erhebung wurde anschließend nach Städten und Unternehmensgröße getrennt. Mikrounternehmen setzen sich dabei aus weniger als fünf und kleine Unternehmen aus fünf bis 20 Arbeitnehmern zusammen. Die Probe unterteilt sich dabei...

Words: 1914 - Pages: 8