Scare tactics are not a good option, partially because they don't work for long, and partially because of the negatives side effects of using them.
The first problem with scare tactics is that they stop working quickly, if they work at all, if the threat doesn't lead to actions. If one attempts to scare a child with a non-existent threat, but every time the threat is made nothing happens, they child will quickly learn to ignore it. They won't be afraid of it, and instead they lesson they will be learning is that their parents say things that don't come true; which is a bad lesson to teach a child.
You probably have witnessed a clear example of this in your peers actually. Many parents will tell a kid that they will face a punishment if they continue doing (or refusing to do) a course of action. If you watch some children will respond quickly to such a threat, while other's won't respond at all. What's the difference? Good parents act on the threatened punishment if the kid doesn't do as told, while other's don't. The kids who's parents threaten punishment, but don't consistently follow through on it, have learned that threats of punishment can be safely ignored because nothing will come out of it. The same is true with other scare tactics, if the theratened scare doesn't consistently result in the threatened scare happening if the child action's continue the child learns the threat is meaningless.
As I already stated scare tactics can also affect the trust, and communicaiton, between parent and child. If the child learns their parent is consistently willing to make false claims when trying to scare them into behaving how can the child be confident other claims the parent's make are true?
The biggest issue with trust comes when you try to warn a child about legitmant risks. If a kid has learned that you make up scare threats to get her to do something then why should she believe any of them are real? Why should she believe that if she doesn't look both ways when crossing the street she could be hurt? Why should she believe that playing with matches is dangerous? Why should she worry about unsafe sex, your threat of her getting pregnant was likely just to scare her into not having sex after all (you would be shocked how many teenages I spoke to honestly believe that pregnancy/STD are largely scare tactics from parents to scare them into being virgins and not really that scary).
It also encourages a lifestyle of fear...to a degree. Honestly I'm not sure that it will likely lead to much more fear, but only becaues I suspect the children will learn to ignore the threats rather then taking them seriously so the child will not learn to be afraid specifically. It still is best to focus on positive outcomes then the bad things that can happen, and there is still some potential harm here.
And of course there is the fact that using scare tactics means not using other, better, proven techniques like time out. Time out are wonderful consistent approaches to punishment that a parent should teach a child about early, they are your best way of discplining a child in most cases!
Having said all that lets talk about the difference between 'scare tactics' and warning about legitimate risks. It's entirely appropriate to tell a child "you shouldn't do that because you may get hurt", or otherwise warn of realistic negative outcomes of an action. This helps a child to reason through cause and effect of actions and learn to judge risks. Exaggerating risks to scare a child into compliance is bad, explaining actual risks is not. It's best to not spend you life focusing on only the negative potential outcomes of actions, but it's good to teach a child to recognize legitimant risks.
Of course if you constantly protect a child from a risk they may have a harder time understanding it. For this reason I sometimes let kids experience the side effect of taking a risk, if the risk is minor enough. If I see a child doing something that could result in a mildly upsetting consequence, like fall over but not being seriously hurt, or an easily replaced toy being broken etc, I may warn a child of the risk, explain why it is a risk, then let the child decide if he/she want's to continue with the activity. Sometimes that means they get hurt, or lose a toy, or have something else negative happen. If it does I'll comfort them and make them feel better, but I'll also try to remind them that is why I warned them, and unfortunately if you do the thing they were doing sometimes the consequence is something bad happening.
It may seem harsh to allow a child to come to harm, even minor harm, but I view it as a learning opportunity. It allows the child to see that his/her actions have consequences, and when I warn them of something it's because there is a real risk and not just a scare tactic. It also helps children learn the cause and effect of their actions, and even learn to judge risk taking. A child may end up deciding they are willing to keep doing something fun even if there is a slight chance of falling and scrapping their knee because it's fun enough to warrant the risk, while another child may not consider getting hurt worth the fun. Different children have different views, what matters is that I'm helping the child to learn to judge the cause and affect and measure what risks they consider worth taking and which are too dangerous. And of course it's giving them a choice, and encouraging children to make choices is always good for them.
My point being you should not try to scare children into behaving, as they will figure this out and stop trusting you. You can, and should, inform children of legitimate risks and encourage them to reason through rather an action is worth taking a risk over. The difference is that you should never exaggerate a risk or try to force a child's behavior with it, if you want a child to behave in a specific manner you should use more proven forms of discipline like time out.