Estimating these few model parameters from observations on real data lets us adapt the model such that unforeseen data may be well mimicked by the model at the macrolevel. On the macrolevel we focus on structural properties of collections in community plat- forms, namely characteristic distributions of the size of collections of information items. We empirically show that our model repro- duces these characteristics more accurately than alternative models for describing effects of attracting attention, such as preferential attachment.
My approach is a set-theoretic approach, one that is sensitive to interaction effects. Gender, specifically, is more than an individual characteristic and can be mode- led as a group membership effect using quantitative mod- eling techniques. I argue that an individual’s structural po- sition in a highly gendered arena shapes and constrains an individual’s opportunities. In day-to-day interactions this becomes obvious. We know that there is a qualitative difference between certain categories of people (e.g., white/non-white, Turk/German, mothers/non-mothers). My model combines individual micro-level data with macro-level data to consider these qualitative differences. These data demonstrate that additional analyses, consid- ering the overlapping effects of gender, race, and other commonly considered individual characteristics as so- cially organizing principles, are warranted.
Bertozzi, F., & Bonoli, G. (2009). Measuring flexicurity at the macrolevel - conceptual and data availability challenges. (Working Papers on the Reconciliation of Work and Welfare in Europe, REC-WP 10/2009). Edinburgh: University of Edinburgh, Publication and Dissemination Centre (PUDISCwowe). https://nbn-resolving.org/urn:nbn:de:0168- ssoar-198184
5.3.1 Education and societal collectivism as independent variables
The measurement of education is crucial to address the hypotheses established in chapter 5.2. The variable education was only briefly addressed in chapter 4.3.3 and will now be explained in greater detail. To capture individuals’ education, the analyses rely on years of formal education, which is the most widely recognized and used indicator for individuals’ level of general education (Bates, 1990; Blanchflower et al., 2001; Davidsson and Honig, 2003; Florin et al., 2003; Parker and van Praag, 2006). Furthermore, years of education was established as a relevant predictor for full-time and part-time entrepreneurial activity (Folta et al., 2010; Raffiee and Feng, 2014; see also chapter 4). To obtain this variable, earlier research utilizing the Euromonitor data set was followed (Adam-Müller et al., 2015; Block et al., 2013b). The Euromonitor data set captures the respondents’ age at the end of full-time education (European Commission, 2012). Following previous research (Adam-Müller et al., 2015; Block et al., 2013b) 6 years were subtracted from this figure, the typical starting age of compulsory education (UIS, 2014). Moreover, in accordance with prior research, a lower bound of 9 and an upper bound of 19 years of full-time education was introduced (Adam- Müller et al., 2015; Block et al., 2013b). The lower bound relates to the fact that all countries in the sample require a level of compulsory education of at least 9 years (Barro and Lee, 2013; UIS, 2014). The upper bound of 19 years relates to the maximum number of years in full-timeeducationtypicallyneededtoobtainamaster’sdegree or similar. Since education is central to this chapter, and the operationalization following prior research is not straight forward, two different forms of operationalization of education were also used with similar results (compare chapter 5.4.2)
The author set up a simplistic agent-based model where agents learn with reinforcement observing an incomplete set of variables. The model is employed to generate an artificial dataset that is used to estimate standard macro econometric models. The author shows that the results are qualitatively indistinguishable (in terms of the signs and significances of the coefficients and impulse-responses) from the results obtained with a dataset that emerges in a genuinely rational system.
Obviously, these findings are model-dependent and as such serve little purpose other than to demonstrate there are noticeable differences between the dynamics emerging in the different systems. But are these differences sufficient to distinguish between rational and boundedly rational worlds by estimating the standard macro econometric models? We examine this issue in the next sub-sections by using the artificial datasets to estimate such models.
We adapt a conceptual model used in previous research on incidence of late-stage cancer to describe the cancer screening utilization climate . We estimate an ecolo- gical model, aggregating across individuals in their coun- ties of residence to generate county-level compositional factors reflecting the personal characteristics of the Fee-for-Service (FFS) Medicare population such as age, race or ethnicity, low income or disability status, distance to closest provider, and recent experience moving to a new residence. For the empirical models, we define these compositional factors separately for women in the FFS Medicare population 2003-2005, for use in the BC screen- ing model. For the CRC model, the compositional variables are defined for all FFS beneficiaries during 2001-2005 a . In both BC and CRC models, we include compositional fac- tors and other county-level variables representing context- ual factors related to both supply and demand. We have a single cross section of 3,133 counties, to which we add a second level to the model by including state-level factors describing nurse practitioner laws, shortages of medical doctors ( MDs), prevalence of MediGap insurance among persons aged 65+, state insurance market competition, and two insurance regulatory factors.
All non-dummy variables in our regressions enter in logs so that coefficients can be interpreted as elasticities. The main advantage of this approach is that it allows variation from countries of different sizes and with completely different magnitudes of language learning, migration, and trade flows to drive the results of our model. An estimation in levels would suffer from considerable heteroskedasticity and the results would necessarily be driven by a small number of countries that send a large number of migrants to Germany or trade with Germany a lot. While immigration from or trade with these countries may be economically more relevant because of its magnitude, our main interest is in identifying the mechanisms that drive language learning more generally and, thus, in using variation from as many countries as possible to identify our coefficients. Additionally, an estimation in levels would require that we specify to which extent institutes in cities of different sizes are exposed to changes in our country-level explanatory variables. For example, an absolute change in immigration from France should, in absolute terms, have a larger effect on exam participation in Paris than on exam participation in Nancy. Paris is a larger city and the institute there has a larger catchment area. A log-log estimation does away with this concern, because it assumes that both institutes experience the same relative rather than absolute change in exam participation as a result of a relative increase in migration. 6
6. Productivity growth, through structural change, is the basis for long-term increas- es in GDP, and therefore of life expectancy. However, the transition of dominance from one sector to another (agriculture, manufacturing, services, etc.) involves important shifts in the nature of work available to national populations. Thus, in order for techno- logical change to succeed, some portion of the working population must shift from per- forming work at the lower level of technology, to that of a higher level. However, there are some anomalies here. First of all, the shift to a higher level of technology by a firm will often involve the dismissal of employees working at the less sophisticated level and the hiring of new and younger employees at the more sophisticated level. Those em- ployees who were dismissed are often relatively highly paid skilled workers, who are no longer needed by the firm, and, if they wish to continue to be employed must do so at a firm that will hire them at lower wages and lower occupational skill levels in the services sector (typically retailing, restaurant employment, and transportation. Thus, the restruc- turing process initially involves an increase in unemployment followed by reemploy- ment at a less-skilled and well-paid level. One important alternative to this radical shift toward downward mobility on the part of formerly moderately skilled, middle-aged em- ployees is to shift into self-employment i.e., entrepreneurship at the small business level. Of course, yet another, and very popular, alternative on the part of middle-aged former employees is to drop out of the labour force entirely on the grounds of retire- ment or even illness.
well known theoretical result (Phillips and Moon, 1999) is that when large panels are available, i.e. under the (n; T ) ! 1 case, the fact that n ! 1 entails that a long-run average relationship between two nonstationary panel vectors exists even when the single units do not cointegrate. On the other hand, with …xed n, Granger (1993) considers a model where each equation is a cointegration relationship with one explanatory variable, and …nds that a necessary and su¢ cient condition for cointegration to be maintained after aggregation is that the number of stochastic common trends that generate the nonstationary variables is equal to one. The pres- ence of a greater number of common trends therefore leads to a spurious regression after aggregation. Gonzalo (1993) bases his analysis on a more complex multivariate model and derives a su¢ cient condition for cointegration to hold after aggregation. The conditions laid out by Granger (1993) and Gonzalo (1993) are very restrictive; however, the existence of cointegration at macrolevel is a well established result. Hence the need for a test that is capable of checking whether cointegration holds after aggregation or not.
The same problem appears at mezzolevel. When analysing firms’ growth in the EU 27 countries for 2000 to 2003, Oberhofer (2012) estimated the industry growth total manufacturing value added. He concluded that domestic demand fluctuations created detectable heterogeneity in the reaction among several different firm cohorts, while the adjustments to the European industry recoveries and recessions were homogeneous. In other words, when domestic level fluctuations were considered, company level fluctuation patterns were very far from those at the industry level. Jovanovic and Rousseau (2014) translated the development of Tobin’s Q as firm level cycles and measured co-movement of investments with that measure. They showed that investment may be cyclical for newly established firms while it is counter-cyclical for older ones, highlighting that individual characteristics like age might play a role in how fluctuations influence certain activities often even aggregated at the national (macro) level. They even use an aggregate Q for the industry level performance measure. Bachmann and Bayer (2014) also investigate the connection of investment with business cycles. In their finding, the investment is pro-cyclical, while productivity, output, and employment growth have counter-cyclical dispersions. This means that the latter measures show a completely different business trend when tracked compared with the pattern of investments. They also call attention to the importance of how firms are chosen when mezzo and macrolevel measures are quantified: for example, we may end up with very different charts depending on whether we consider only firms with ongoing operations or all of them.
Second, in addition to presenting a new empirical pattern regarding the electoral effects of fiscal consolidations in emerging economies, we test a mechanism through which adjustments have political effects, mainly concentrating on the way the consolidation is executed both in terms of its specific design (composition) and the timing of the fiscal adjustment process, in particular with respect to the state of the business cycle (recessions, expansions) surrounding fiscal decisions. These features have usually been recognized as important determinants of the success of fiscal consolidations in the economic literature, but their political consequences remain understudied. Finally, we present data from an original survey experiment that provides a micro foundation for the finding that governments are more likely to lose elections if they increase taxes in Latin America. Our survey is devised to gauge voter preferences on specific policy instruments (taxes, spending) using a vignette design that randomly exposes respondents to different economic scenarios (recessions, expansions). The vignettes allow us to measure voter policy choices directly, allowing for a comparison between voter preferences and the actual policy choices observed at the macrolevel.
The analysis suggests that the economic effects of multilateral liberalization are rather limited for Brazil. Accordingly, poverty would remain largely unaffected by such reforms. In contrast, a full liberalization scenario implies quite substantial welfare gains that are concentrated among some of the poorest groups of the country, in particular those in agriculture. This scenario is also most interesting from a methodological viewpoint, as it highlights the benefits of a behavioral micro-simulation. Under full liberalization, the rural poor benefit more than pro- portionately, a result driven—on the macrolevel—by an export boom in agriculture and agricultural processing industries, growing labor demand and associated higher wages. However, following full liberalization, a larger number of workers remain in agriculture compared to the baseline scenario. Given that moving out of agriculture may substantially improve the income situation of a household, one may expect full liberalization to weaken poverty reduction, an expectation supported by the obser- vation that moving households are on average poorer than those remaining in agriculture (for example because they are landless). However, this is not the case, as the gain in agricultural incomes more than compensates the reduced benefits from lower migration flows (for example because they are better educated than those who stay in agriculture).
The present paper adopts a VAR-GARCH approach to model the dynamic linkages be- tween both the mean and the variance of macro news and commodity returns. This is in contrast to the vast majority of earlier contributions, which only examined level eﬀects. Analysing simultaneously the interactions between the first and second moments sheds new light on the issues of interest. The layout of the paper is the following. Section 2 outlines the econometric modelling approach. Section 3 describes the data and presents the empirical findings. Section 4 summarises the main findings and oﬀers some concluding remarks.
Fewer studies have examined the eﬀects of macro news on commodity prices. Despite not being financial assets, the latter have been shown to be aﬀected by variables such as interest rates (Frankel, 2008) and the US dollar exchange rate, both of which are known to respond to news announcements. Frankel and Hardouvelis (1985) provide evidence of a statistically significant response to US money supply announcements; eﬀects of macro news on various commodity prices are also found by Cai et al. (2001), Hess et al. (2008), Kilian and Vega (2008); commodities futures prices have been reported to be aﬀected as well (Barnhart, 1989; Ghura, 1990). Roache and Rossi (2010) in particular show that they are influenced by the surprise element in macro news, with evidence of a pro-cyclical bias after controlling for the eﬀects of the US dollar, the only exception being gold, which reacts counter-cyclically given its role as a safe heaven and store of value, and is more sensitive to bad news and higher uncertainty. Unlike most other authors, typically using OLS, they estimates a GARCH(1,1) model given the evidence of time variation and clustering of volatilities (Cai et al., 2001, is another of the few papers using a GARCH framework, specifically to examine the impact of news on gold futures prices).
also adversely a¤ect creditor economies through wealth losses on external asset positions and trade channels.
In the other direction, international …nancial integration may also have positive e¤ects on output growth. First, cross-border net …nancial ‡ows may accelerate convergence, if capital is reallocated from capital-abundant economies to capital-scarce economies. Sec- ond, international risk sharing (through trade in equity-type instruments) may facilitate the selection of higher-return, higher-risk projects that generate faster (if more volatile) output growth (Obstfeld 1994). Moreover, for a given level of output volatility, international risk sharing reduces the welfare costs of volatility by insulating wealth and consumption from domestic production shocks.
interest lies primarily in obtaining dynamic estimates of macro-financial spillovers 8 . This
will allow us to investigate how the strength and structure of spillovers has varied over time and whether the changes appear to relate to specific economic and financial events. To achieve this we use a standard rolling window estimation approach in which the parameters of the MF-VAR, the FEVD arrays and the connectedness measures are re- estimated each time the window is rolled forward. A window length of 60 months is employed, since it appears to offer a good balance between providing a sufficient sample size to estimate the parameters of the underlying MF-VAR to an appropriate level of accuracy, and allowing dynamics of connectedness to be captured. The study of business cycle connectedness by Diebold and Yılmaz (2012b) also employed the same rolling window length with a monthly dataset of industrial production series. We have however checked the robustness of our results to reasonable changes in the window length, with these results available upon request.
“HEX 316” - A pavilion designed to exhibit 12 parametrically designed Jewelry pieces designed by students during the 2017-year end campus festival. It is designed concentrically on site with a cantilevered entrance. The parametric structure is made out of 316 hexagonal shaped cells with 3 different types of modules each serving a different function. The cells were generated with patterns to create internally and externally faces. The openings gradually increase toward the cantilevered structure in order to create openings at eye level and reduce the weight on the cantilevered structure. The cells are built from 6mm thick corrugated cardboard. All the cells consist of two components - the facing and the framing. Cardboard is cut, scored and folded to create these components and bolted together.
in the 1980s. First and most firmly, we find that the transmission of macro and financial shocks across borders is largely asymmetric, going from the US to the euro area. Previous literature hints at this asymmetry, but does not fully model the bidirectional spillovers, or does it for only one policy or aspect. For instance, Jarocinski (2019) shows using a SVAR, that Fed monetary policy has much stronger effects on ECB’s monetary policy while euro area’s has negligible impact on the US. Second, we find that the intensity of transmissions across borders increased over time. However, since the Great Recession this positive trend has been reversed, and the transmission of US shocks has been weakened. This could be a result from the weakened dominance of the US economy globally, or due to the protectionism that followed the financial crisis of 2007-08. Third, we find a negative relation in the transmission between euro area shocks and US financial conditions. One could say that adverse (favourable) shocks in euro area developments could be beneficial (damaging) for the US financial conditions. However, this pattern has also been weakened following the near financial meltdown in the US in 2008. Previous works (Berg and Vu (2019), Gourinchas et al. (2019), Jarocinski (2019), Giorgiadis (2016)) have advocated for a dominant position of the US in the international financial, monetary or macroeconomic sphere. However, as far as we are aware, this is the first study to formally establish this in a structural empirical model with (i) full bidirectional spillovers between two of the largest global economies, (ii) along macroeconomic and financial dimensions simultaneously and (iii) covering a relatively large time span that allows for long-term interpretations.
- FinMaP- Policy Letter No.2 - P a g e | 4 Figure 1: japanes Credit Network (square are banks, circle firms, colors represent the leverage level, red agents are riskier)
Both simulated and empirical data show that growth of credit and connectivity increases the probability of having a crisis. Indeed, when credit increases usually also the level of leverage for both firms and banks augments, thus their individual riskiness grows. Moreover, credit growth is often associated with the raise of credit network connectivity. When agents are riskier and more connected among them even a small local shock as the failure of a relatively small firm may be easily amplified and diffused affecting the whole economic system. Indeed, the variation of network connectivity may be used to increase the effectiveness of credit variation as early warning indicators of crises. Early warning measure effectiveness derives from their capacity of reducing the trade-off between false alarms and hits. A good early warning measure is the one that is able to anticipate crises occurrence (hits) but at the same time does not generate too many false alarms. The ROC curve represents the trade-off between false alarm and hits (see Figure 2) and the larger is that area below the curve (the so-called AUROC) the higher the effectiveness of the indicator.