No abstract

To gain a better understanding of the impact that COVID-19 is having on food systems and rural livelihoods in the region, researchers in the Agricultural Policy Research in Africa (APRA) Programme of the Future Agricultures Consortium (FAC) are conducting a rolling series of telephone-based household surveys and key informant interviews in selected study locations across multiple countries. This report presents results from the first round of that research in seven countries – Ethiopia, Ghana, Kenya, Malawi, Nigeria, Tanzania and Zimbabwe – from interviews conducted in June-July 2020.1 APRA will monitor the situation as the pandemic unfolds through further rounds of data collection and analysis in late 2020 and early 2021.

We investigate the relationship between the corporate income tax burden and firm size in South Africa using a panel dataset from companies' tax returns. We find that medium-sized companies are experiencing the lowest effective tax rate, while the smallest companies are facing the highest effective tax rate.

While it is widely accepted that Africa is experiencing a youth employment crisis, the nature of the crisis is disputed. In relation to rural youth, the crisis is variously framed in term so unemployment, underemployment, missing jobs, a lack of decent work, waithood and mixed or diverse livelihoods; with each framing pointing toward a different response. We look more closely at how young people engage with the labour market using a small, high-frequency dataset that includes activities of 233 individuals aged 18–24 years in rural areas of Ghana and Uganda. Specifically, we describe four dimensions of their work (its nature, frequency, steadiness and amount), analyse relationships between these dimensions, and link them with characteristics of the study participants. We conclude that in the early phases of livelihood building non-domestic work activities of young people are multi-faceted, context and seasonally specific, and highly gendered. This reflects, in part, different priorities given to education, domestic work, childbearing and social relations relative to economic activities. This study highlights the need for a better understanding of the various factors—including individual priorities—that come into play in the early phases of livelihood building, and their implications for when and how young people engage with non-domestic work.

No abstract

About UONGOZI Institute 'Uongozi' means leadership in Kiswahili, and inspiring and strengthening leadership is the core purpose of our organization. UONGOZI Institute is dedicated to supporting African leaders to attain sustainable development for their nations and for Africa. This is done through the provision of high-quality executive education (leadership competencies), facilitation of policy dialogues, action-oriented research and technical assistance for public and private institutions.

Convolutional Neural Networks are extensively used in a wide range of applications, commonly including computer vision tasks like image and video classification, recognition and segmentation. Recent research results demonstrate that multilayer (deep) network involving mono-dimensional convolutions and dilation can be effectively used in time series and sequences classification and segmentation, as well as in tasks involving sequence modelling. These structures, commonly referred to as Temporal Convolutional Networks (TCNs), have been demonstrated to consistently outperform Recurrent Neural Networks in terms of accuracy and training time [1]. While FPGA-based inference accelerators for classic CNNs are widespread, literature is lacking in a quantitative evaluation of their usability on inference for TCN models. In this paper we present such an evaluation, considering a CNN accelerator with specific features supporting TCN kernels as a reference and a set of state-ofthe-art TCNs as benchmark. Experimental results show that, during TCN execution, operational intensity can be critical for the overall performance. We propose a convolution scheduling based on batch processing that can boost efficiency up to 96% of theoretical peak performance. Overall we can achieve up to 111,8 GOPS/s and a power efficiency of 33,9 GOPS/s/W on an Ultrascale+ ZU3EG (up to 10x speedup and 3x power efficiency improvement with respect to pure software implementation).

I evaluate the impact of BNDES disbursements on Brazilian commercial banks' disbursement using balance-sheet data for the period of 2002-2016. Using dynamic panel data techniques, I find BNDES disbursement for both investment in innovation and fixed capital investments crowded-in commercial banks' disbursement. Further, the results obtained considering the distribution before and after 2008, suggest the beginning of the crowding-in impact together with the countercyclical role adopted by the bank at the beginning of the financial crisis.