Key Empirical Finding Research Articles

A machine learning-based framework for forecasting sales of new products with short life cycles using deep neural networks

Demand forecasting is becoming increasingly important as firms launch new products with short life cycles more frequently. This paper provides a framework based on state-of-the-art techniques that enables firms to use quantitative methods to forecast sales of newly launched, short-lived products that are similar to previous products when there is limited availability of historical sales data for the new product. In addition to exploiting historical data using time-series clustering, we perform data augmentation to generate sufficient sales data and consider two quantitative cluster assignment methods. We apply one traditional statistical (ARIMAX) and three machine learning methods based on deep neural networks (DNNs) – long short-term memory, gated recurrent units, and convolutional neural networks. Using two large data sets, we investigate the forecasting methods’ comparative performance and, for the larger data set, show that clustering generally results in substantially lower forecast errors. Our key empirical finding is that simple ARIMAX considerably outperforms the more advanced DNNs, with mean absolute errors up to 21%–24% lower. However, when adding Gaussian white noise in our robustness analysis, we find that ARIMAX’s performance deteriorates dramatically, whereas the considered DNNs display robust performance. Our results provide insights for practitioners on when to use advanced deep learning methods and when to use traditional methods.

Investigating the employee–customer relationship in a utilitarian context

ABSTRACT This research, drawing on dyadic data, is one of the first to assess the significance of employee satisfaction in driving customer satisfaction and service quality within a utilitarian context. Multi-level structural equation modelling is employed to analyse 974 customers directly matched to 95 employees across 15 retail grocery stores. Our key empirical finding is that within a utilitarian context, and using a multi-level approach, employee satisfaction does not have a significant direct effect on either customer satisfaction or customer service quality perceptions. Perceptions of price competitiveness assume a more significant role in both customer satisfaction and service quality perceptions than employee satisfaction. Utilitarian store managers should focus their efforts on appropriate service strategies where employee satisfaction still plays a key role.

Implications of Political Stability and Financial innovations for Sectoral Economic Growth of Pakistan

The economic status of a country plays an important role in the lives of the citizens of both developed and developing countries. With the rapid increase in population, every country is trying to find new ways to boost its growth rate so that it may elicit the maximum number of people from poverty and can compare other countries in the form of improvement. The main objective of the study is to empirically investigate the impact of political stability and financial innovations on economic growth. Furthermore, the study also analyzes the impact of subsectors (Agricultural, Industry, and Services) of the economy on growth. A time-series co-integration autoregressive distributed lag (ARDL) model is used to investigate the general to specific sector's growth. The key empirical finding shows that political stability and financial development has a positive and significant impact on economic growth as well as its sub-sectors in the long-run. Trade liberalization has a positive impact on economic growth but most surprising results have been witnessed in agriculture growth for Pakistan.

Measuring and assessing the effects of collaborative innovation in crime prevention

This study reports the findings from an interactive research project in which academics and practitioners worked closely together in designing a new, criteria-based assessment tool that enables local municipalities to measure the degree of collaboration, innovation and crime-preventive effect in publicly financed projects intended to keep at-risk youth out of criminal activities. The assessment tool also offers a much-needed opportunity for researchers to study the extent to which cross-boundary collaboration may spur the development of innovative solutions, which in turn may help to prevent youth crime. The key empirical finding is that collaboration has a strong association with public innovation, which in turn has a strong association with the ability of local projects to help prevent crime. The result makes an important contribution to the expanding field of public innovation research in which quantitative studies that combine process evaluation and impact studies are extremely rare.

Similarity between deviant peers: Developmental trends in influence and selection*

AbstractThe association between an adolescent's own behavior and that of his or her peers remains a key empirical finding in the study of delinquency, and this similarity is often explained in criminology by invoking processes of social influence and homophily. Adolescence is a period of rapid change for both individuals and their surroundings, however, and influence and homophily are often discussed without attending to their development over time. In the current study, I employ longitudinal social network models to estimate social influence and homophily related to alcohol and cigarette use and to determine whether there is change in the strength of these processes. Furthermore, I test whether the broader social environment conditions these processes and their observed changes. The results from this study indicate that although influence from peers decreases from early to mid‐adolescence, homophily on drinking and smoking increases. There is also evidence that school‐level opportunities and expectations affect social influence and homophily but do not account for the average changes observed in these processes.

Forecasting New Product Life Cycle Curves: Practical Approach and Empirical Analysis

We present an approach to forecast customer orders of ready-to-launch new products that are similar to past products. The approach fits product life cycle (PLC) curves to historical customer order data, clusters the curves of similar products, and uses the representative curve of the new product’s cluster to generate its forecast. We propose three families of curves to fit the PLC: bass diffusion curves, polynomial curves, and simple piecewise-linear curves (triangles and trapezoids). Using a large data set of customer orders for 4,037,826 units of 170 Dell computer products sold over three and a half years, we compare goodness of fit and complexity for these families of curves. Fourth-order polynomial curves provide the best in-sample fit with piecewise-linear curves a close second. Using a trapezoidal fit, we find that the PLCs in our data have very short maturity stages; more than 20% have no maturity stage and are best fit by a triangle. The fitted PLC curves of similar products are clustered either by known product characteristics or by data-driven clustering. Our key empirical finding is that, for our large data set, data-driven clustering of simple triangles and trapezoids, which are simple to estimate and explain, perform best for forecasting. Our conservative out-of-sample forecast evaluation, using data-driven clustering of triangles and trapezoids, results in mean absolute errors approximately 2%–3% below Dell’s forecasts. We also apply our method to a second data set of a smaller company and find consistent results. The online appendix is available at https://doi.org/10.1287/msom.2017.0691 .

Children's out-of-home leisure activities: changes during the last decade in Norway

ABSTRACT This study, based on two nationally representative samples of children aged 6–12 years and their parents from 2005 and 2013–2014, explores changes in children's play, through changes in participation in three different out-of-home leisure activities during the last decade in Norway as well as the interactions between these activities. The study finds that there has been an increase in the share of children who participate in organized leisure activities; that they start to participate at a younger age; and that the intensity of participation in organized leisure activities has increased. During the same period, there has been a decrease both in the extent of unsupervised play outdoors and in the degree to which children visit friends’ houses. A key empirical finding is that there is a strong positive association between participation in organized leisure activities and, respectively, the extent of unsupervised outdoor play and visits to friends.

Modeling the migration of cultural goods: a macro-economic study of domestic Chinese cinema

ABSTRACTThis article investigates the migration of cultural goods such as movies in the burgeoning commercial exhibition marketplace of China. Based on prior research on the movie industry and theatrical releases, this empirical study models the dynamic relationships between cinema attendance and the number of cinema screens in response to consumer mindset shifts in terms of predetermined and exogenous variables. These relationships are established and estimated by applying the Vector Auto-Regression (VAR) and Seemingly Unrelated Regressions (SUR) techniques. The regression analyses are executed over 96 observations spanning from May 2005 to April 2013. A key empirical finding is that cultural goods such as movies are affected by a variety of market and economic factors. Another significant finding is that income and admission price will not significantly affect the Chinese cinema-goers’ consumption behaviors, whereas they may alter the market expectation of the exhibitors. In recent years, the prevalence of Internet options has further promoted a more competitive market of choice, quality, and prices. This is an area worthy of further scholarly investigation.

Profitable Nonprofits? Reward-Based Crowdfunding in the Czech Republic

AbstractThis paper deals with crowdfunding as a new and alternative mechanism of nonprofit funding which has thus far attracted little scholarly attention in the Czech context. It focuses on the factors that affect a campaign’s overfunding rate and determines whether these factors are consistent among different forms of project creators. A comprehensive analysis of reward-based crowdfunding in the Czech Republic was conducted based on the data from 617 projects using the Czech crowdfunding platform Hithit that were categorized according to the status of the project creator. Using binary logistic regression, a statistic estimation on an overall sample and on a sample of purely nonprofit projects was executed in order to observe if the outcomes diff er. Th e key empirical finding is that NGOs tend to raise fewer additional funds than other forms of project creators. Except for certain specific factors, the effects of the examined factors were consistent across all samples. Th is new and innovative approach to resource insufficiency by using alternative funding sources presents an important and unexplored research gap in the (post-) transitive context of nonprofit sector studies, enabling a view of policy implications for Czech NGOs.

Trading on Talent: Human Capital and Firm Performance

How is skilled human capital reflected in firm performance? By directly observing the monthly career migration patterns of 37 million employees of US public companies, along with their education, demographics, and skills, we explore firm-level skill premia. Our key empirical finding is that, contrary to the individual-level patterns documented by the labor economics literature, technical and social skillsets negatively forecast both financial and operational performance at the firm level. We explore several potential mechanisms for this finding. Social skillsets display counter-cyclical performance, suggesting that their negative premia reflect their risk profiles. Meanwhile, negative premia on technical skillsets show patterns consistent with over-exuberance regarding contemporaneous popular technologies: IT and Mobile Networks skillsets carry negative premia in the early 2000s, while Data Analysis, Software Engineering, and Web Development display negative premia during the 2010s.

Components of party polarization in the US House of Representatives

We specify the level of polarization in a two-party legislature as an explicit function of three factors: (1) the ideological heterogeneity of district median voters, (2) the distance between candidates of different parties in the same or ideologically comparable districts, and (3) partisan bias in choosing between candidates equidistant from the median voter. Our key empirical finding, reinforced by two alternative methods of calculation, is that, while changes in each factor have contributed to the present day extremely high level polarization in the US House of Representatives, at least 80% of the growth in that polarization from 1956 through 2008 can be attributed to a dramatic increase in the second of these factors: party differentiation at the district level.

Sex Trafficking and the State: Applying Domestic Abuse Interventions to Serve Victims of Sex Trafficking

Advocacy and scholarship addressing sex trafficking as a human rights issue has become a transnational effort, but there has been less attention to sub-national efficacy. Through analyzing progressive justice system responses to domestic violence in Duluth, Minnesota that have been adopted worldwide, this paper demonstrates how to effectively apply these local advances in order to address sex trafficking on a global scale. This paper makes a theoretical contribution to understanding the intersections between domestic abuse and sex trafficking. A key empirical finding is that a coordinated community response (including the justice system and women’s organizations) is crucial for advancing domestic abuse training, monitoring, and legislation—and this coordination can also be productively utilized for improving responsiveness to victims of sex trafficking across a diverse range of socio-legal and economic contexts.

Estimating (Markov-Switching) VAR Models without Gibbs Sampling: A Sequential Monte Carlo Approach

Vector autoregressions with Markov-switching parameters (MS-VARs) fit the data better than do their constant-parameter predecessors. However, Bayesian inference for MS-VARs with existing algorithms remains challenging. For our first contribution, we show that Sequential Monte Carlo (SMC) estimators accurately estimate Bayesian MS-VAR posteriors. Relative to multi-step, model-specific MCMC routines, SMC has the advantages of generality, parallelizability, and freedom from reliance on particular analytical relationships between prior and likelihood. For our second contribution, we use SMC's flexibility to demonstrate that the choice of prior drives the key empirical finding of Sims, Waggoner, and Zha (2008) as much as does the data.

An investigation of the causal relations between exchange rates and interest rate differentials using wavelets

This paper uses wavelet analysis to investigate causality between the spot exchange rate and the nominal interest rate differential for seven country pairs, which includes Sweden. Impulse response functions are also utilized to examine the signs of how one of these variables affects the other over time. One key empirical finding from the causality tests is that there is strengthening evidence of the nominal interest rate differential Granger causing the exchange rate as the wavelet time scale increases. When considering impulse responses on how the interest rate differential affects the exchange rate, there appears to be some evidence of more negative relationships at the shorter time scales (i.e. an increase in the Swedish interest rate compared to that of another country is associated with a lower Swedish krona price of the other country's currency) and more positive relationships at the longer time scales.

Institutionalizations compared: Implementating the EU's 2008 wine reform

In 2008 the European Union adopted an apparently radical reform of its wine policy. However, whether this policy change actually reinstitutionalizes the regulation of this industry is dependent on how it is implemented and represented by both commercial operators and regulators located at national and intra-national scales. Using interview and documentary data generated on this process in France, Spain and Romania, this article sets out to compare and explain the differentiated institutionalization of the reform that has actually occurred to date. Its key empirical finding is that reinstitutionalization is dependent on the degree to which change has been legitimated within such regions. More precisely, durable change or maintenance of the status quo stems from how conflicts over different parts of the EU's reform have been framed and debated locally. This finding has two wider implications for institutionalist theories of political change. First, constructivist accounts of the framing of collective and public problems need to be brought to the fore. Second, at least when studying contemporary Europe, the question of scale and its institutional effects should be systematically built into research strategies and methodologies.

A modelling process of short-term interest rate risk management for the South African commercial banking sector

This study focuses on banking book interest rate risk (IRR) management, more specifically short-term IRR management (SIRR). This type of risk is partly induced by the inflation targeting policy of the South African Reserve Bank (SARB). As a result, inflation leads to an uncertain interest rate cycle and a period of uncertain interest rate levels as it relates to lending and borrowing activities in the South African commercial banking sector. This study highlights what causes short-term interest rate risk and how the banks may forecast and manage the SIRR with reference to the inflation targeting policy. The banking industry manages a high volume of fund transactions and portfolios of investments. The banks are intricately involved in the financial markets and are therefore exposed to a large number of risk factors. A sound banking system is an important prerequisite for a country’s future economic development. One key empirical finding of this research is that 50 per cent of the South African banks agree that loans that cannot undergo immediate rate adjustments are exposed to the repo-rate adjustment after the Monetary Policy Committee (MPC) meeting. Banks surveyed see the need for the development of a short-term interest rate risk (SIRR) management process to better control such repo-rate risk. The next key empirical finding is that interest rate risk is still managed via traditional repricing gap and sensitivity analysis which is not ideal for risk management due to inherent weaknesses (such as not quantifying capital risk exposure). This agrees with the Pricewaterhousecoopers Balance Sheet Management benchmark survey

Europeanisation in new member and candidate states

The Europeanisation of candidate countries and new members is a rather recent research area that has grown strongly since the early 2000s. Research in this area has developed primarily in the context of the EU’s eastern enlargement. A small number of theoretically informed book-length studies of the EU’s influence on the Central and Eastern European candidate countries have provided a generalisable conceptual framework for this research area, drawing on the debate between rationalist institutionalist and constructivist institutionalist approaches in International Relations and Comparative Politics. This framework makes these studies highly compatible with analyses of the Europeanisation of member states, with which they also share one key empirical finding, namely that the impact of the EU on candidate countries is differential across countries and issue areas. At the same time, the theoretical implications of these findings appear more clear-cut than in the case of the Europeanisation of member states: rationalist institutionalism, with its focus on the external incentives underpinning EU conditionality and the material costs incurred by domestic veto players, appears well-suited to explaining variation in the patterns of Europeanisation in candidate countries. A very recent development within this research agenda is the focus on the Europeanisation of new member states. While the study of the EU’s impact during the early years of membership was hitherto primarily a subfield of analyses of the Europeanisation of member states, it has now become an extension of studies of candidate countries by analysing the impact of accession on the dynamics of pre-accession Europeanisation and how durable and distinctive the patterns of candidate Europeanisation are in the post-accession stage.

The Green Solow model

We argue that a key empirical finding in environmental economics—the Environmental Kuznets Curve (EKC)—and the core model of modern macroeconomics—the Solow model—are intimately related. Once we amend the Solow model to incorporate technological progress in abatement, the EKC is a necessary by product of convergence to a sustainable growth path. We explain why current methods for estimating an EKC are likely to fail; provide an alternative empirical method directly tied to our theory; and estimate our model on carbon emissions from 173 countries over the 1960–1998 period.

UK stock price effects of permanent and transitory shocks

This paper examines the dynamic relationship between stock prices and dividends using a structural cointegrated vector autoregression. The approach adopted fully identifies the system without imposing arbitrary restrictions and decomposes innovations into permanent and transitory components. Prior research indicates that transitory price shocks could lead to stock price predictability. Our key new empirical finding is that permanent dividend shocks could also lead to aggregate stock price predictability in the UK.

The Determinants of Labor Market Earnings in a Small Caribbean Island: The Case of Dominica

This paper uses the Labor Force Survey (LFS) from 1999 to investigate the labor market earnings determination process in the small Eastern Caribbean island of Dominica. Given the interval coded nature of the earnings data reported for the LFS, an interval regression model estimated by maximum likelihood techniques is used. A key empirical finding is that the Dominica labor market places a relatively high valuation on formally acquired post-primary human capital assets. We also find that the ceteris paribus public sector pay premium is relatively large and suggests public sector workers are securing a high rent through employment in this sector. A gender pay gap of approximately 20 percent is detected and there is also evidence of a sizeable ethnic pay disadvantage for male members of the island’s indigenous population.