Have environmental regulations restrained FDI in China? New evidence from a panel threshold model.

This paper aims to investigate the effect of environmental regulations on inward foreign direct investment in China. For this purpose, a panel threshold model was constructed to assess the threshold effects of environmental regulations on the influx of foreign direct investments (FDI) . The findings indicate that, under the influence of human capital, the impact of environmental regulations on FDI in China was characterized by a V-shaped curve, indicating an initial inhibitory effect followed by a subsequent increase. A plausible explanation is that specific pollution-generating FDI must withdraw from China because of stringent environmental regulations before human capital reaches a certain threshold level. Meanwhile, impaired by the adverse selection effect, some cleaner-production FDI cannot easily enter China. As a result, environmental regulations in this stage have an inhibitory effect on FDI in China. However, part of the pollution-generating FDI is converted into cleaner production after the human capital level reaches the threshold limit. Further, due to the positive selection effect, additional cleaner-production FDI can also enter China from different destinations. At this stage, environmental regulations boost overall FDI entering China.

Generalized Matrix Local Low Rank Representation by Random Projection and Submatrix Propagation.

Matrix low rank approximation is an effective method to reduce or eliminate the statistical redundancy of its components. Compared with the traditional global low rank methods such as singular value decomposition (SVD), local low rank approximation methods are more advantageous to uncover interpretable data structures when clear duality exists between the rows and columns of the matrix. Local low rank approximation is equivalent to low rank submatrix detection. Unfortunately, existing local low rank approximation methods can detect only submatrices of specific mean structure, which may miss a substantial amount of true and interesting patterns. In this work, we develop a novel matrix computational framework called RPSP (Random Probing based submatrix Propagation) that provides an effective solution for the general matrix local low rank representation problem. RPSP detects local low rank patterns that grow from small submatrices of low rank property, which are determined by a random projection approach. RPSP is supported by theories of random projection. Experiments on synthetic data demonstrate that RPSP outperforms all state-of-the-art methods, with the capacity to robustly and correctly identify the low rank matrices when the pattern has a similar mean as the background, background noise is heteroscedastic and multiple patterns present in the data. On real-world datasets, RPSP also demonstrates its effectiveness in identifying interpretable local low rank matrices.