Simultaneous perturbation methods for adaptive labor staffing in service systems

We consider the problem of optimizing the workforce of a service system. Adapting the staffing levels in such systems is non-trivial due to large variations in workload and the large number of system parameters do not allow for a brute force search. Further, because these parameters change on a weekly basis, the optimization should not take longer than a few hours. Our aim is to find the optimum staffing levels from a discrete high-dimensional parameter set, that minimizes the long run average of the single-stage cost function, while adhering to the constraints relating to queue stability and service-level agreement (SLA) compliance. The single-stage cost function balances the conflicting objectives of utilizing workers better and attaining the target SLAs. We formulate this problem as a constrained parameterized Markov cost process parameterized by the (discrete) staffing levels. We propose novel simultaneous perturbation stochastic approximation (SPSA)-based algorithms for solving the above problem. The algorithms include both first-order as well as second-order methods and incorporate SPSA-based gradient/Hessian estimates for primal descent, while performing dual ascent for the Lagrange multipliers. Both algorithms are online and update the staffing levels in an incremental fashion. Further, they involve a certain generalized smooth projection operator, which is essential to project the continuous-valued worker parameter tuned by our algorithms onto the discrete set. The smoothness is necessary to ensure that the underlying transition dynamics of the constrained Markov cost process is itself smooth (as a function of the continuous-valued parameter): a critical requirement to prove the convergence of both algorithms. We validate our algorithms via performance simulations based on data from five real-life service systems. For the sake of comparison, we also implement a scatter search based algorithm using state-of-the-art optimization tool-kit OptQuest. From the experiments, we observe that both our algorithms converge empirically and consistently outperform OptQuest in most of the settings considered. This finding coupled with the computational advantage of our algorithms make them amenable for adaptive labor staffing in real-life service systems.

Unemployment rates and population changes in Spain

Published as an article in: Journal of Applied Economics, 2004, vol. VII, pages 47-76.

Export performancecof Indian spices in the WTO regime: a disaggregated analysis

Export has assumed an important place in the development of any country and considered as the engine of economic growth. India requires huge amount of foreign exchange for its essential import and for achieving rapid growth. Millions of job opportunities have to be created to utilise the youth for nation building. Even though the country has different sources of foreign exchange, export earning is the safe way of obtaining it in the long run. Export of high valued traditional products not only gives foreign exchange, but also employment to large number of people. Spices are the traditional products of India whose production process is highly intensive in semi and unskilled labour, and high domestic and foreign market prices compared to other traditional products. The new world trade scenario with the establishment of WTO has affected India’s spices export considerably. The study examines the export performance of Indian spices in the WTO regime taking the export of major spices from 1985 to 2013 using the growth of export, trend and instability in growth rate, changes in the composition and direction of spices, export performance ratio and the prospects of spices in earning foreign exchange during the WTO period and Pre-WTO period. The analysis reveals that the overall performance of Indian spices exports during the WTO regime are satisfactory. Export volume and value increased much during this period. But the decrease in market share of spices export during the WTO period reflects that, the favourable conditions in the international market are not exploited by India. High Revealed Comparative Advantage (RCA) and Value Elasticity (EV) of major spices amidst the low export shares shows that export performance of Indian spices during the WTO regime was not mainly affected by external demand factors as suggested by Ragnar Nurkse in his Demand Deficiency Thesis, but because of internal supply factors as suggested in Supply Deficiency Thesis, (supported by K.S Dhinsha, Dacosta, Goddamwar,etc.). But the fluctuations of export during the recession and prosperity periods show that external demand is also a determinant of Indian spices export. From this one can conclude that both the domestic supply factors and foreign demand factors influence the export performance of Indian spices. The long term performance of Indian spices exports are mainly influenced by domestic supply factors as suggested by Supply Deficiency Thesis and short term performance is mostly influenced by external demand factors as suggested by Demand Deficiency Thesis.

¿A quiénes afecta el desempleo? análisis de la tasa de incidencia en Colombia

El análisis de la dimensión de riesgo en el estudio del desempleo, complementa el estudio de aspectos como su tipología, la importancia de la intermediación laboral o la vulnerabilidad de ciertos grupos. En este sentido, la tasa de incidencia representa un indicador compuesto que tiene en cuenta el volumen de desempleados y la persistencia en este estado mediante la incorporación de la duración media del desempleo. El análisis de la tasa de incidencia permite caracterizar de manera más completa a quienes tienen una mayor probabilidad de entrar en el desempleo o permanecer en esta situación. . Se encuentra que en Colombia existen diferencias significativas entre la tasa de desempleo y la tasa de incidencia, lo que implica que la situación del mercado de trabajo no sólo se explica por el efecto que los choques económicos tienen sobre la composición de la oferta y demanda de trabajo sino también por los fenómenos de duración en los diferentes estados laborales. Estos eventos se pueden considerar igualmente importantes para explicar la dinámica de corto y mediano plazo del mercado laboral.

Más allá de la infraestructura: el impacto de las bibliotecas públicas en la calidad de la educación

La literatura sobre la calidad de la educación ha prestado poca atención al papel que tienen las bibliotecas públicas dentro de los determinantes del desempeño educativo. Las bibliotecas públicas son activos externos al colegio y al hogar del estudiante, pero hacen parte del entorno social que les rodea. La puesta en marcha a finales de 2001 de tres bibliotecas de gran tamaño en Bogotá, conocidas como megabibliotecas, nos permite analizar el impacto de estas iniciativas sobre la calidad de la educación en los colegios aledaños. Dicho impacto se daría a través de mecanismos adicionales a la simple reducción de costos al acceso a la información: las bibliotecas renovaron el espacio público mediante la generación de espacios agradables y amigables hacia la educación, además ofrecen regularmente actividades lúdicas dirigidas a las habitantes del sector. Aprovechando la distancia del plantel educativo a la biblioteca como una aproximación al costo de acceso a la misma, utilizando para ello Diferencia en Diferencias junto a la descomposición Blinder Oaxaca. Encontramos que las mismas parecen no tener un impacto significativo sobre el desempeño académico general en los exámenes oficiales SABER 11 durante los años posteriores a su implementación. Se recomienda analizar programas específicos que aprovechen las bibliotecas para actividades escolares y otras posibles variables de impacto como actitudes hacia el estudio y aspiraciones a la educación superior.

¿A quiénes afecta el desempleo? análisis de la tasa de incidencia en Colombia

El análisis de la dimensión de riesgo en el estudio del desempleo, complementa el estudio de aspectos como su tipología, la importancia de la intermediación laboral o la vulnerabilidad de ciertos grupos. En este sentido, la tasa de incidencia representa un indicador compuesto que tiene en cuenta el volumen de desempleados y la persistencia en este estado mediante la incorporación de la duración media del desempleo. El análisis de la tasa de incidencia permite caracterizar de manera más completa a quienes tienen una mayor probabilidad de entrar en el desempleo o permanecer en esta situación. Se encuentra que en Colombia existen diferencias significativas entre la tasa de desempleo y la tasa de incidencia, lo que implica que la situación del mercado de trabajo no sólo se explica por el efecto que los choques económicos tienen sobre la composición de la oferta y demanda de trabajo sino también por los fenómenos de duración en los diferentes estados laborales. Estos eventos se pueden considerar igualmente importantes para explicar la dinámica de corto y mediano plazo del mercado laboral.

Asistencia a Educación Preescolar y Desempeño Escolar en el Largo Plazo, Evidencia para Colombia

Reconociendo la importancia que tienen las inversiones durante los primeros años del ciclo de vida, en este documento se estima el efecto de la asistencia a preescolar sobre el desempeño escolar de largo plazo, medido a través del puntaje en las áreas de lenguaje y matemáticas de la prueba Saber 11 en Colombia. Para ello, se realiza la aproximación empírica a través de la metodología de variables instrumentales. Los resultados indican que la asistencia a un año adicional de educación preescolar reduce en 0.121 desviaciones estándar el puntaje obtenido en el área de lenguaje. No obstante, se identifica un efecto diferenciado a partir de variables que dan cuenta del estatus socioeconómico de los individuos.

Shrinking the posterior: A note on the Nerlovian model

Diebold and Lamb (1997) argue that since the long-run elasticity of supply derived from the Nerlovian model entails a ratio of random variables, it is without moments. They propose minimum expected loss estimation to correct this problem but in so-doing ignore the fact that a non white-noise-error is implicit in the model. We show that, as a consequence the estimator is biased and demonstrate that Bayesian estimation which fully accounts for the error structure is preferable.

Public expenditures taxation and welfare measurement

this article addresses the welfare and macroeconomics effects of fiscal policy in a frarnework where govemment chooses tax rates and the distribution of revenues between consumption and investment. We construct and simulate a model where public consumption affects individuaIs' utility and public capital is an argument of the production function. The simulations suggest that by simply reallocating expenditures from consumption to investment, the govemment can increase the equilibrium leveIs of capital stock, hours worked, output and labor productivity. Funhennore, we 'show that the magnitude and direction of the long run impact of fiscal policy depends on the size of the elasticity of output to public capital. If this parameter is high enough, it may be the case that capital stock, within limits, increases with tax rates.

Mercado de capitais e financiamento privado para o desenvolvimento

This research argues that Brazil should create conditions for long run financing of development. The end of the inflationary process is just a first step. A second step is the development of an institutional framework to offer good alternatives to investors that have long run targets, such as pension funds etc. Particularly, in the Brazilian market predominates the trade of preferred shares instead of common shares that give more prerrogatives for investors that plan to hold these shares on long term basis. This attitude turns the market more volatile and in this way corporations lose the chance of financing their projects with large amounts of capital and have to rely instead on more debt which weakens their financial strenght.

A Common-feature approach for testing present-value restrictions with financial data

It is well known that cointegration between the level of two variables (labeled Yt and yt in this paper) is a necessary condition to assess the empirical validity of a present-value model (PV and PVM, respectively, hereafter) linking them. The work on cointegration has been so prevalent that it is often overlooked that another necessary condition for the PVM to hold is that the forecast error entailed by the model is orthogonal to the past. The basis of this result is the use of rational expectations in forecasting future values of variables in the PVM. If this condition fails, the present-value equation will not be valid, since it will contain an additional term capturing the (non-zero) conditional expected value of future error terms. Our article has a few novel contributions, but two stand out. First, in testing for PVMs, we advise to split the restrictions implied by PV relationships into orthogonality conditions (or reduced rank restrictions) before additional tests on the value of parameters. We show that PV relationships entail a weak-form common feature relationship as in Hecq, Palm, and Urbain (2006) and in Athanasopoulos, Guillén, Issler and Vahid (2011) and also a polynomial serial-correlation common feature relationship as in Cubadda and Hecq (2001), which represent restrictions on dynamic models which allow several tests for the existence of PV relationships to be used. Because these relationships occur mostly with nancial data, we propose tests based on generalized method of moment (GMM) estimates, where it is straightforward to propose robust tests in the presence of heteroskedasticity. We also propose a robust Wald test developed to investigate the presence of reduced rank models. Their performance is evaluated in a Monte-Carlo exercise. Second, in the context of asset pricing, we propose applying a permanent-transitory (PT) decomposition based on Beveridge and Nelson (1981), which focus on extracting the long-run component of asset prices, a key concept in modern nancial theory as discussed in Alvarez and Jermann (2005), Hansen and Scheinkman (2009), and Nieuwerburgh, Lustig, Verdelhan (2010). Here again we can exploit the results developed in the common cycle literature to easily extract permament and transitory components under both long and also short-run restrictions. The techniques discussed herein are applied to long span annual data on long- and short-term interest rates and on price and dividend for the U.S. economy. In both applications we do not reject the existence of a common cyclical feature vector linking these two series. Extracting the long-run component shows the usefulness of our approach and highlights the presence of asset-pricing bubbles.

Análise das relações de longo prazo entre a posição internacional de investimentos, o efeito Balassa-Samuelson e a taxa de câmbio real: testes de cointegração

Este trabalho tem a finalidade de analisar as evidências de relações de longo prazo entre a taxa de câmbio real (“RER”), a posição internacional de investimentos (“NFA”) e o efeito Balassa-Samuelson (“PREL”) em um grupo de 28 países, grupo este que inclui países em diferentes estágios de desenvolvimento. A metodologia utilizada foi a de testes de cointegração. Os testes aplicados foram desenvolvidos por Bierens (1997), teste não paramétrico, e por Saikkonen e Lütkepohl (2000a, b, c), teste que consiste em primeiro estimar um termo determinístico. Evidências de cointegração são constatadas, em ambos os testes, na maioria dos países estudados. Entretanto, houve diferenças relevantes entre os resultados encontrados através dos dois testes aplicados. Estas diferenças entre os resultados, bem como alguns casos especiais de países que não demonstraram evidências de cointegração, requerem análises mais aprofundadas sobre o comportamento de longo prazo das três variáveis estudadas.

O impacto dos gastos com publicidade nas vendas das firmas: avaliação empírica.

Este trabalho tem por objetivo estimar um modelo empírico para relacionar os gastos em publicidade com a receita das firmas, de forma a servir como ferramenta de tomada de decisão, para isso vamos fazer um estudo de caso da indústria de telecomunicações. A Indústria de comunicação (publicidade) no Brasil, segundo dados do IBGE de 2008, é responsável por 4% do PIB, gerando receitas da ordem 115 bilhões de reais. Com 113 mil empresas que geram 711 mil empregos, ocupam 866 mil pessoas e pagam 11,8 bilhões em salários e encargos. No entanto, a maioria dos gestores de marketing declara não ter instrumentos para medir o impacto de suas ações no resultado das empresas. O modelo empírico será estimado tendo como base dados mensais dos serviços de ligações de longa distância nacional da Embratel para o período de janeiro de 2009 até dezembro de 2011. As informações quase sempre não disponíveis, só puderam ser usadas devido ao compromisso de confidencialidade. A partir de técnicas de cointegração, foi calculada a elasticidade de longo prazo da receita em relação aos gastos com publicidade e ao preço, assim com as respectivas velocidades de ajustamento aos desvios de curto prazo. Os resultados sugerem que a receita responde positivamente às variações dos gastos em publicidade, embora o percentual seja relativamente baixo, através do teorema de Dorfman-Steiner conseguimos indicar que o ponto ótimo da relação entre gastos com publicidade e a receita seria de aproximadamente 20%, respeitadas as limitações do modelo.

Ensaios sobre política monetária

Esta tese é composta de três ensaios a respeito de política monetária. O primeiro ensaio aborda o canal em que as crises financeiras aumentam a ineficiência alocativa nos países emergentes. O segundo ensaio trata do grau de não-neutralidade da moeda no Brasil de acordo com o modelo de Golosov e Lucas (2007). O terceiro ensaio estima a inclinação da hazard function da precifi cação para o Brasil pela metodologia de Finite Mixture Model.

Addressing important econometric issues on how to construct theoretical based exchange rate misalignment estimates

Exchange rate misalignment assessment is becoming more relevant in recent period particularly after the nancial crisis of 2008. There are di erent methodologies to address real exchange rate misalignment. The real exchange misalignment is de ned as the di erence between actual real e ective exchange rate and some equilibrium norm. Di erent norms are available in the literature. Our paper aims to contribute to the literature by showing that Behavioral Equilibrium Exchange Rate approach (BEER) adopted by Clark & MacDonald (1999), Ubide et al. (1999), Faruqee (1994), Aguirre & Calderón (2005) and Kubota (2009) among others can be improved in two following manners. The rst one consists of jointly modeling real e ective exchange rate, trade balance and net foreign asset position. The second one has to do with the possibility of explicitly testing over identifying restrictions implied by economic theory and allowing the analyst to show that these restrictions are not falsi ed by the empirical evidence. If the economic based identifying restrictions are not rejected it is also possible to decompose exchange rate misalignment in two pieces, one related to long run fundamentals of exchange rate and the other related to external account imbalances. We also discuss some necessary conditions that should be satis ed for disrcarding trade balance information without compromising exchange rate misalignment assessment. A statistical (but not a theoretical) identifying strategy for calculating exchange rate misalignment is also discussed. We illustrate the advantages of our approach by analyzing the Brazilian case. We show that the traditional approach disregard important information of external accounts equilibrium for this economy.