Watching this is a reminder of how much better the best dog trainers are today than they were 25 years ago.
This is Vicki Hearne, who was a very good writer and dog trainer. Here she is in 1991 trying to rehabilitate Bandit, a heavy-boned Pit Bull that was slated to be killed for biting people.
Vicki Hearne was at the top of the heap in 1991, and criticism is cheap. That said, I would observe two things: 1) Vicki barely has control of the dog because the dog is about as powerful as she is and she just has a slip chain collar on the dog, and not a prong collar, and; 2) The dog has been brought too close to the cows too soon.
If this dog had been put on a prong collar, and things moved a bit slower, there would be less drama. But perhaps drama was what this film producer required? That could be it, as the push from TV and film crews did not start yesterday, did it?
I post this clip, not to criticize what it being done here, but to note that when folks are asked what to do to stop a specific behavior, they seem to jog backwards very fast. Suddenly it's all very complicated and they have to know "why" the behavior is being done. But guess what? Most of the time it doesn't matter why the behavior was being done. Bandit has some prey drive. Maybe he's just curious about those cows, but when he approaches and they run, the code will explode. So how do we get Bandit to be less curious about cows? How do we get him to not chase cows and look at cows as being about as interesting as a rock or a fence?
And what if it isn't a cow, but a dog? What would you do then? How would you get a leash-reactive dog to chill out and become phlegmatic? Might it be the same thing for dogs as for cows?
How about a squirrel? A chicken?
Now, to be fair, none of this was as easy 25 years ago as it is today.
Twenty five years ago, the main tool was a long leash, and so the timing of a corrections tended to be poor even if you were half-decent at long-leash handling. So while Vicki could have done much better with a prong collar on this short leash, 25 years ago she would have been stuck with either a long leash on a slip collar, or a thrown object such as a chain, rope, or bumper, or an old-fashioned "buster" e-collar with perhaps 7 levels which would have been quite a nit hotter than the low-stimulation "tap" collars we use today.
Having said that it would have take longer 25 years ago, the basics would have been about the same: you cannot be shy about correcting bad behavior, and you reward calmness and good behavior, and that too is done with good timing. You do not need to "understand" why the dog chases the cow, or hates the other dog, or barks at squirrels. You just have to send a strong NO signal, and reward calmness (and not necessarily with food).
If the dog looks at the cow -- correction. When the dog fully relaxes: a bit of kibble or a gentle scratch behind the ear.
Repeat, repeat, and repeat again as you move closer and closer to the problem. At first the dog will be on a long leash tied to a firm post or stake, but that will not be needed at some point. You can train a dog not to chase cows, or sheep, or chickens. But can it be done with treats alone? Maybe not in a high drive dog.
Which reminds me of the late Greg D. He trained his Patterdale terriers to leave chickens alone by putting them in, at a pretty early age, with an Asil fighing chicken. The bird only had natural spurs, but it was enough, and afterwards the dog never ordered chicken off the menu again. Another friend did much the same thing with a tough old ewe, and the adult border terrier was well broken to sheep after that.
Bottom line: Get on with it. If you have a problem with a self-reinforcing behavior that has to stop, name it and identify it, and move to find an aversive or punishment that is well-timed enough, and disliked enough, that it stops the behavior.