In crowding, flankers impair perception of a target. For example, Vernier offset discrimination deteriorates when the Vernier is flanked by parallel lines. Pooling models explain crowding by averaging of neural activity corresponding to the Vernier and the flankers, thus, reducing signal to noise ratio. However, recently, it was shown that flankers, longer than the Vernier, lead to less crowding than equal length flankers. Adding additional long flankers reduced crowding almost fully- in stark contrast to pooling models which predict just the opposite result. These and other findings clearly show that crowding cannot be explained by local spatial interactions, but global computations are needed. Here, we show that a Wilson-Cowan type model can explain both classical, local and recent, global aspects of crowding. The Wilson-Cowan type model employs end-stopped receptive fields with isotropic excitatory connections and anisotropic lateral inhibitory connections that are reciprocal between adjacent neurons. The key feature of the models is spread of neural activity across similar elements which are eliminated during recurrent inhibition. For example, crowding strength decreases the more long flankers are presented because these similar, long flankers inhibit each other during time consuming processing and, thus, reduce inhibition on the dissimilar Vernier. For equal length flankers, the Vernier is “treated” similarly to a flanker and inhibited. For this reason, and in accordance with psychophysical data, crowding does not vary with the number of equal length flankers.